T he 2002/3 SAPS Annual Report is, in many respects, an improvement on previous editions. While it is still more of an advocacy piece than a detached analysis of police progress, the report makes a real attempt to explain the causes behind crime trends, an improvement over the submissions of the last two years. Issue could be taken with many aspects of this discussion, particularly the sections on the impact of HIV/AIDS and vigilantism. But it is encouraging to see a little of the old crime analysis coming back into the picture, especially as the Annual Report has become the only regular public document released by the state on the vital topic of crime and its prevention. The bottom line for the South African public is the release of the crime statistics, and this year the Minister has graciously expanded the information available to the public by running area-level recorded crime figures on the Internet. It is on these figures that the following discussion is based. Murders and attempts Murder is down once again, at least as a rate. The actual number of victims increased ever so slightly, but the trend is clearly downward since 1994. This is reassuring because murder is the one form of violent crime that is not heavily under- reported, and it is thus the most reliable indicator of the real violent crime situation. At 47 murders per 100,000 citizens (about the same as the most dangerous urban area of the United States, Washington, D.C.), the situation is still dire. But we are no longer in top contention for the title of ‘murder capital of the world’. Columbia, our old rival for the honour, has long surpassed us, with 66 murders per 100,000 in 2002. The ‘success ratio’ of murders to attempted murders has shown a distinct decline since 1994 (Figure 1). 45 SA CRIME QUARTERLY No 6 DECEMBER 2003 1 Ted Leggett, Institute for Security Studies ted@iss.co.za THE FACTS BEHIND THE FIGURES Crime statistics 2002/3 The new crime figures indicate that violent crime continues to decline, although violent acquisitive crime (robbery) is on the increase. However, these broad national trends conceal vast regional differences, including the continued growth of crime in the Western Cape. And a close look at some of the new figures suggest that crime recording rules continue to be refined in the post-moratorium period. Figure 1: Success rate of attempted murders % s u cc es sf u l 1994/5 30 35 40 50 55 1995/6 1996/7 1997/8 1998/9 1999/0 2000/1 2001/2 2002/3 This suggests that the increase in recorded attempted murders is a result of an increased rate of reporting by the public, not an increase in real incidence. The alternative is to believe that we are 11% less effective at killing one another today than we were in 1994. But the picture is not uniformly rosy. The number of murders in the Western Cape is at an all time high. Certain policing areas within the Western Cape have been particularly badly afflicted: murder in the East Metropole is up nearly 20% in the last year, up over 60% from 1994/5. Robbery By far the most troubling trend as far as national recorded crime is concerned is the increase in reported robberies. Although we know this is a crime that is heavily under-reported (about 41% of 1997 National Victim Survey respondents who were victims of robbery said they reported it), the increase is so striking that it is difficult to believe that there is not a real increase in incidence. Nationally, the number of aggravated robberies reported is up about 50% since 1994, but in many areas, including the Western Cape, the number of incidents has more than doubled. Unless nearly everyone is now reporting, which would be unusual by international standards, things have genuinely got worse. The national commissioner last year and the minister this year have suggested that an increase in fraudulent reporting is behind this boom, largely tied to insurance fraud involving cell phones. Increases in the number of people insured could result in an increase in reporting. But if these cases are indeed fraudulent, one would expect a corresponding increase in fraud cases. In fact, fraud cases are down sharply since 1999/2000. To their credit, the police have this year started keeping track of robberies that occur in the home as a subset of aggravated robbery, an innovation the ISS has recommended in the past. This crime combines the worst aspects of robbery (the use or threat of violence) with the worst aspects of burglary (violation of the sanctity of the home), producing a crime more serious than either offence considered separately. SA CRIME QUARTERLY No 6 DECEMBER 2003 2 LEGGETT The majority of these crimes occur in Gauteng, which suggests that this is a criminal technique with the potential to be exported from the big city. After dipping to a six-year low in 2001 (162 incidents), cash-in-transit robberies shot up to 421 in 2002. But this is most likely a result of changes in recording practices by the SAPS, because with regard to bank robberies, the opposite trend is observed: 432 incidents recorded in 2001, and only 153 in 2002. Looking at the phenomenon at a more local level, the number of cash-in-transit robberies in Johannesburg policing area went from 16 in the last financial year to 42 in 2002/3, while bank robberies dropped from 56 to eight. Overall, what the SAPS is calling ‘bank related robbery’ dropped from 594 in 2001 to 574 in 2002, down from over 1,000 in 1994. This is good news, since this is not a crime likely to be under-reported, and these improvements could be directly attributed to improvements in security and enforcement. Car thefts and hijackings Like murder and bank robbery, vehicular thefts and robberies are highly reported crimes, due to the fact that cars and trucks are valuable items and are often insured. Declines in these figures should be taken seriously, and incidents of both offences have seen national reductions. The SAPS announced with justifiable pride that national hijacking figures were down by 20%. Hijacking, which is often tied to organised crime syndicates, is precisely the type of crime which the SAPS is well-equipped to deal with on a national level, and this trend could well be the direct result of enforcement efforts. Unfortunately, how this reduction was achieved has not been documented for the public. The most remarkable decline has been not in carjacking, but in truck hijacking. It is probable that this is partly an artefact of changing recording practices (‘is a bakkie a truck?’), but it may partly be due to the fact that truck hijacking is even more likely to be tied to organised crime than carjacking. At national level, the number of truck hijackings dropped from 3,333 in the previous year to less than 1,000 in 2002/3. Truck hijackings are at a national low from the time they were first recorded as a separate crime category, down from a high of over 6,000 in 1998/9. But these aggregate trends conceal stark regional contrasts. Truck hijackings were down by only 4% in the Eastern Cape, contrasted to an 83% reduction in KwaZulu-Natal. And in some areas, like Johannesburg, the decline has been nothing short of precipitous. In 1996/7, the Johannesburg policing area produced about a quarter of all truck hijackings nationally, with over 900 incidents that year. Last year, there were only 51. Even if some of those early ‘trucks’ were Sandton SUVs, this has to be seen as progress. Looking at the crime that concerns most of us - carjacking - the reduction has been significant but more modest: about 8% better than last year in terms of total incidents. According to the official rates, our national chances of being carjacked are now down to about what they were in 1996/7. However, certain areas have experienced remarkable increases. In the Western Cape, carjackings have more than tripled since the first year the SAPS started counting them (1996/7), and are up more than 20% in the last year alone. In the West Metropole, the number of carjackings this year is more than four times that of 1996/7. Gauteng still has by far the highest rate of carjackings, but the risk of victimisation has remained fairly stable in recent years. This provincial trend belies significant local changes, however. While there has been little change in Soweto, the number of incidents in Pretoria have increased from 371 in 1996/7 to 892 last year. This figure is, however, down from 998 Pretoria carjackings in 2001/2. Nationally, car theft has been coming down sharply since a peak in 1998/9. The share of stolen vehicles taken by force (that is, carjackings versus vehicular thefts) has fluctuated between 15% and 17% since 1996/7, before reaching a new low this year at 14%. This is encouraging because it runs counter to the theory that hijackings are a response to increasing levels of vehicular security, and that we can expect SA CRIME QUARTERLY No 6 DECEMBER 2003 3LEGGETT more violent means of acquiring property to follow attempts to protect ourselves. Locally, however, there is considerable variation on this trend. In Gauteng in 1996/7, the share of violent vehicular thefts was 16%. In 2001/2, it was 19%. This year saw a reduction to 18%, but this was due to a decline in truck hijackings, not carjackings. Other crimes Trends in many other offence categories are difficult to discuss, because they are often heavily influenced by reporting rates. Particularly for violent interpersonal crimes like rape and assault, various campaigns and pieces of legislation designed to increase reporting rates should, if successful, lead to increases in recorded cases. It is therefore inappropriate to celebrate the decline in recorded rape cases. This dip could be the result of increasing disenchantment with the criminal justice system rather than a real decline in incidence. But the same holds for overall increases in assault, which could be nothing more than diligent application of the Domestic Violence Act coming into play. It also appears that the rape decrease may, once again, be partly due to shifting recording patterns. Indecent assaults went up by nearly as much as rape went down (Figure 2), so this could be the repackaging of what is nearly the same number of offences. The statement is made in the Annual Report that “everybody should know by now that... at least one out of three rapes is reported to the SAPS”, citing surveys by, among others, the ISS. In fact, everybody, especially the SAPS, should know by now that this is not true. This figure is based on victim surveys, but many women who refuse to report rape to the police might also be hesitant to mention it to some fieldworker who comes to her door, especially if her assailant is standing behind her in the one room shack. While the actual reporting rate remains unknown, it is undoubtedly higher than one in three, in line with underreporting rates found in countries where SA CRIME QUARTERLY No 6 DECEMBER 2003 4 LEGGETT better research has been done in this area. If anything, lack of services and cultural issues should make the rate of rape reporting less here than in better-developed countries. A crime area that also seems to have been subject to definitional re-jigging is kidnapping/abduction. While abduction cases shot up by a quarter, kidnapping cases dropped by almost the same amount, for a minor aggregated decrease. A similar trend is seen at local level. The number of child abuse cases nearly doubled between 2001/2 and 2002/3, a change so dramatic that it is probably also a recording phenomenon. Like rape, this is a massively under-reported crime, and so the change is actually encouraging. The total is still under 5,000 cases nationally, a tiny total when compared to (also under-reported) cases of child sexual abuse. Unless child rape is more common than other forms of abuse, there is considerable scope for improvement in reporting in this area. Yet another pair of offences that may be subject to definitional shifts are commercial and residential burglary. Commercial burglary is down 15% in one year, to its lowest levels since 1994, while residential burglary is up by 5%. Because commercial burglaries are less common than residential ones, the total of these two offence categories is actually stable. Since most commercial premises are more likely to be insured than domestic ones, commercial burglary is a more reliable indicator of real burglary levels, so this year’s figures are good news. Like other under-reported crimes, increases in theft rates should be taken with a grain of salt. The better- developed countries have theft rates far in excess of that of South Africa, due to the fact that people diligently report minor crimes out of a sense of civic duty, trust the police to take such cases seriously, and are required to report for insurance purposes. As South Africa becomes better developed, we should expect even more thefts to be recorded every year. Regional risks While the SAPS seems adamant about denying the public access to station-level crime statistics, the very figures we need to protect ourselves and to lobby for resources, they have done us the service of identifying the police areas and station areas that produce the highest numbers of offences. Unfortunately, these lists are actually quite unfair. By referring to numbers of offences committed rather than using a rate based on the population size, the larger policing areas and station areas are unfairly prejudiced. The populations of policing areas vary widely, from under 100,000 (Namaqualand) to nearly three million (Umtata). For example, Marico in the North West makes it to the top 11 policing areas for murder, but this is not because the average citizen of the area is as much at risk as one living in the Johannesburg policing area. Marico has one of the largest populations of any station area, projected to be over 1.8 million in 1999. Johannesburg, on the other hand, was estimated to have less than 900,000 residents during the same period. So even if your chances of being murdered were twice as bad in Johannesburg as Marico, the number of dead bodies produced would be the same. Figure 2: Recorded sexual assaults 2000/1 52,000 54,000 50,000 48,000 46,000 56,000 58,000 60,000 62,000 64,000 N o . o f in ci d en ts Rape Indecent Assault 5 2 ,8 7 5 6 ,6 5 2 2001/2 5 4 ,2 9 3 7 ,6 8 3 2002/3 5 2 ,4 2 5 8 ,8 1 5 SA CRIME QUARTERLY No 6 DECEMBER 2003 5 The same is true for the station area comparisons. Station areas can serve communities of a few thousand or a few hundred thousand. Thus, mega stations like Moroka, Mitchell’s Plain and Khayelitsha make it onto almost every list. In the end, this information is useless if you want to know where you are most at risk. What is more interesting is when small station areas pop up from time to time. Although it has a tiny resident population (just over 20,000), Johannesburg Central remains among the top stations listed for both murder and aggravated robbery, despite crowing about the success of CCTV in that area. On a larger scale, it is clear from the preceding discussion that the Western Cape has both the worst and the fastest growing crime situation in the country. The reasons behind this are the subject of current ISS research, but it is clear that the Cape must be the focus of intensive attention by the SAPS if it is not to swing entirely out of control. Good, but not good enough For the most part, the crime rates are out of the control of the SAPS, and so should not be used as a benchmark of police performance. When considered in light of changing reporting trends, however, they do give an indication that South Africa is beginning to move beyond the violence of its tumultuous past. Certain trends, like the decline in bank-related robberies and hijackings, do reflect well on the work of the police, but these phenomena are best understood at a local level of analysis. What is still lacking are station-level crime statistics. The present operational strategy of the police, the National Crime Combating Strategy, is based on the idea that enforcement in specific high-crime areas will stabilise crime nationally, but in order to evaluate this claim, station-level figures are essential. Until these are released to the public, the impression will remain that claims of crime reduction are simply smoke and mirrors. LEGGETT