In South Africa, housing for the poor takes many forms. International observers may regard famous townships like Soweto as emblematic of South Africa’s remarkable income inequality, failing to recognise the diversity of lifestyles contained in these sprawling areas. Locals are more likely to call to mind our massive shack lands, where thousands of people live in ad hoc structures without access to the amenities of the 20th century, or rural settlements where subsistence agriculture is still the order of the day. But South Africa also has the kind of housing for the poor that persists in the first world, despite decades-long campaigns to eliminate it – inner-city slums.

While Hillbrow police station is responsible for a diverse area containing many types of housing, it is most notorious for the low-income tenements that most people refer to as ‘Hillbrow’. Once the residential annex to the city centre, capital flight long ago sent the area down the slippery slope towards the marginal. For many years, the area tread water above the sleaze line, being both safe and risqué enough to attract the young and adventurous, as well as many African travellers on a budget. But post-1994, the area has finally crossed beyond the pale for all but the most desperate.

Today, it is said to have one of the highest population densities of any area in the country: an estimated 100,000 people in an area of 10.28 square kilometres.¹ The actual level of crowding is probably much worse, however, because a substantial portion of the residents are squatters and undocumented migrants, both groups likely to evade official head counts. Many of the buildings, owned by absentee landlords, teeter on the edge of collapse. Their interiors range from quaintly tatty to positively post-apocalyptic.

Among South Africans, Hillbrow is renowned for two things: immigrants and crime. It is arguably the most feared neighbourhood in the country, edging out rivals like Mannenburg and Mitchell’s Plain in the Western Cape. Whether this reputation is deserved is debatable, as is the relationship between the immigrant population and the crime problem, and this monograph represents one step in an attempt to probe the validity of popular prejudice.

Hillbrow is also well known for being the epicentre of drug activity in the country, and both retailers and consumers will travel considerable distances to purchase their drugs at source. With drugs, particularly crack cocaine, comes prostitution, and many residential hotels in the area are, essentially, high-rise brothels. These residential hotels provide the anonymity and flexibility required for all sorts of criminal commerce. Drug addicts don’t always have cash but they will always find something to trade, so Hillbrow has become a prime conduit for stolen property. Drug couriers need travel documents, and this leads to a broadening web of forgery and fraud. And, as is the case in most communities of this sort, the marginalised prey on their own, first and most consistently.
`
`Drugs also mean corruption, here as everywhere. The demand is such that the cost of bribes can easily be passed on to the consumer, and the profits provide plenty of fat for market shocks. Policing drug areas poses a major challenge to traditional law enforcement techniques, and demands that as much effort be spent watching the watchers as patrolling the streets.
`
The area administratively defined as Hillbrow is small, and most people use the name to refer to a larger area, embracing parts of Berea and Joubert Park. For the local residents, ‘Hillbrow’ has more to do with the character of the place than arbitrary boundaries. The Hillbrow police station area is more expansive still, involving very different sorts of neighbourhoods, such as Killarney, a largely white suburb to the north.

Immediately to the south of this station area is another police sation area, prosaically dubbed Johannesburg Central. The station is housed in a building of considerable apartheid-era infamy, then known as John Vorster Square. The station area contains some of the most important public buildings in the country in its 13.19 square kilometres, including the Reserve Bank, the Gauteng parliament and executive offices, the ANC national headquarters, and many others. This means that marches aimed at protesting provincial and national policy tend to become the responsibility of Johannesburg Central. The area also contains the headquarters of many of the largest banks in the country. Despite all this, it is scarcely in better repair than Hillbrow, with many large buildings standing vacant, and squatters and tenants residing in crumbling office blocks. The crime situation, tied more to alcohol than drugs, is comparable nonetheless.

Crimes recorded by the police

The number of crimes recorded by the police in these two station areas has varied quite a bit over the years, but some trends are clear. In Hillbrow, robberies with firearms nearly doubled between 1996 and 2000, and, as will be discussed, armed robbery is one of the biggest problems confronting the area. Assault with intent to cause grievous bodily harm has also sharply increased. On the other hand, theft of vehicles and residential burglary have just as radically declined over the same period (Figure 1.1). A similar trend is seen in Johannesburg Central (Figure 1.2). This suggests a shift from stealthy acquisitive crime to violent crime, acquisitive and otherwise. The reasons for this shift may be linked to increasing drug use and will be discussed in Chapter 6.

Figure 1.1: Incidents of crime recorded by police in Hillbrow




Figure 1.2: Incidents of crime recorded by police in Johannesburg Central



Comparing the crime scenes in these two areas to other parts of the country and the world poses something of a problem. Crime comparisons are usually done in terms of crime rates, most often expressed as crimes per 100,000 residents per annum. The problem is, as mentioned above, it is very difficult to come up with reliable population estimates for these areas. The official 1999 population totals, extrapolated from the 1996 Census data, are 96,862 for Hillbrow and 22,707 for Johannesburg Central.² Given the number of illegal immigrants in the area, many residing in blocks not zoned for residence, these figures may be serious underestimates.

In addition, both of these policing areas experience large commuter inflows. Surveys done at the major nodal interchanges show massive numbers of people transiting the area. The Metro Mall transportation node experiences an estimated 150,000 to 200,000 commuters per working day, while the Joubert Park taxi rank has an additional 100,000. Some of these people work in the area while some go on to destinations in Sandton and Randburg, but all are potential crime victims for as long as they are in the jurisdictional boundaries.³

A quick look at the crime figures highlights how difficult it is to calculate crime statistics based on the official figures. Johannesburg Central experienced 219 murders in 1999, or over 1,000 murders per 100,000 residents. This is entirely off the scale of international comparison, and suggests that 1% of the population is murdered each year. If this rate of homicide were indeed the case, everyone in the area would be dead by the end of the century.

Looking at Figures 1.1 and 1.2, it becomes clear that, with the exception of residential burglary, the number of crime incidents in the two areas is fairly comparable. Johannesburg Central experienced about 30% fewer serious assaults and about 30% more vehicle thefts than did Hillbrow, while robberies were about the same. But, according to Census projections, the population of Johannesburg Central is just over a fifth of that of Hillbrow. That would mean the crime rate in Johannesburg Central is five times worse than that in its more notorious northern neighbour, a picture that does not gel with popular experience.

Policing

The notion that Johannesburg Central presents a greater crime problem than Hillbrow seems to be supported by the national police authorities, judging by the level of police station staffing. According to station figures gathered in the ISS’ time use study of the area in 2002, Johannesburg Central has over two-and-a half times the staff of Hillbrow, with 863 employees to Hillbrow’s 327. This is despite the fact that, according to the Census, Johannesburg Central has a quarter of the resident population and a land area just over 20% larger. This means that there are 296 members of the public to each Hillbrow station member and only 26 citizens per member of Johannesburg Central. However, about 15% of Johannesburg Central’s staff is assigned to guarding the local courts, a task that Hillbrow does not have to do, and Johannesburg Central also has a higher general administrative burden. Setting all this aside, Johannesburg Central still fields about 300 uniformed crime prevention members, while the equivalent division in Hillbrow has just over 60 staff (Figures 1.3).

Figure 1.3: Comparative police staffing levels



The number of crimes recorded at each station is similar, with Hillbrow generally recording more serious violent crime and Johannesburg Central more property crime (except for burglary) (Figure 1.4).

Figure 1.4: Comparative crime incidents recorded by police, 2000



One likely reason for this division of resources is the need for a large uniformed standby force in Johannesburg Central in case of a major public march. Johannesburg Central contains a number of important buildings housing a number of important people, and remains very much in the public eye. Hillbrow, on the other hand, is populated by a far more marginal community, without the collective voice to demand more protection.

Closed circuit television

The policing approaches of the two stations are fairly traditional, in keeping with the operational strategies of the national office, with one notable exception. The Johannesburg Central policing area is one of the few areas of the country to have extensive closed circuit television (CCTV) coverage, a programme initiated by Business Against Crime and the SAPS in April 2000. As of late 2001, there were 75 cameras covering large areas of the city centre, and there were plans to expand this number to 360 by June 2002. By December 2002, however, there were only 184 cameras active. The system in place has the capacity to handle 4000 cameras, which, while it would be enormously expensive, would surely provide blanket coverage of the entire poliing area.

Various claims have been made as to the crime reducing qualities of the presence of cameras, both here and in Cape Town. Since complete crime statistics for the SAPS station area level for 2001 and 2002 have not yet been released, these claims are difficult to verify. Looking at the crime data for the first half of 2001, from six months to a year after the launch of the system, the broader policing area of Johannesburg remains first in the nation for most serious crime types. Thus, even if specific crime reductions were claimed for the areas where the cameras were focused, this crime may very well have been simply displaced to elsewhere in the policing area. Furthermore, looking specifically at the Johannesburg Central station area data for 2000 and the first half of 2001, no overall improvement is seen.

But using these figures to evaluate Johannesburg CCTV is probably unfair. It is very difficult to evaluate claims of crime prevention on the basis of official crime figures because these figures are inherently unreliable. This is because the police can only record those crimes which they witness or which are reported to them by the public. If the police do a good job, both the number of crimes they spot on the streets and the willingness of the public to report crimes should increase. The present survey shows that the majority of serious crimes in inner Johannesburg currently go unrecorded because the public does not report them, so there is plenty of room for improvement in this area. Furthermore, systems like CCTV should assist in detecting more crime, and so could cause an increase in the number of crimes recorded, not the contrary.

There are reasons to be sceptical of the crime preventative effect of CCTV, however. In the inner city context in particular, preventing crime through street surveillance would be limited in its potential, as most of the population at any given time is situated vertically, in high rise apartment and office blocks. Studies in the US have shown that street-focused enforcement has driven criminal activity such as drug dealing indoors, increasing its negative impact on the local community.⁴ And even with regard to street coverage, the cameras face serious challenges. Many of these buildings have concrete overhangs that shield the sidewalk from view. Hawkers use umbrellas to shade their stands and trees shade many areas. Coverage is therefore incomplete, and it would not be difficult for habitual criminals to adjust their behaviour accordingly.

Perhaps the fairest way to gauge the effectiveness of the CCTV system is to look at it from a more appropriate set of police performance indicators: responses to crime. We would expect that with the aid of millions of rands of high tech equipment, police members assigned to the CCTV response teams would show much higher arrest rates than those without this advantage. This proves not to be the case.

As of the time of going to press, there were 52 SAPS members assigned full time to CCTV duty. During the first 11 months of the year 2002, these members combined made between two and 27 arrests each month, with an average of 11 arrests monthly. That means each member made an average of about one arrest every six months, in a station area that suffers over 20,000 crimes a year. These members are backed by an even more numerous civilian monitoring staff and equipment estimated to cost over R100 million. ‘Operating costs’ are estimated at R2.5 million a year, enough to employ an army of 70 starting constables. The simple purchase cost of a single camera is R180,000, which would pay a constable to stand on that same corner for five years.

Thus, CCTV does not seem to be a cost-effective method of increasing the number of criminals taken off the streets. Whether it has effectively reduced crime in the areas that it covers and in the station area as a whole as a result of its deterrent effect is another question, and one that would require a detailed and independent study, with access to up-to-date crime statistics. Before public funds are invested in very costly projects of this sort, there should be some evidence that such a system actually works in the South African context.

The need for a victim survey

In order to evaluate whether the overall approach taken by the two stations is proving effective in terms of real victimisation levels and public opinion, there is no getting around the need for regular public surveys. As noted above, the official crime figures are not reliable indicators of the real crime situation, so the only way to find out how much the people of inner Johannesburg are suffering is to go to their homes and ask them. In addition, a good deal of attention has been paid in the SAPS to improving service delivery independent of crime prevention efforts, and this work also needs continual reassessment. The rest of this monograph is dedicated to laying out the responses of the over 1,300 people polled in both station areas in mid-2002.