In January 2012, four senior commissioned officers from the South African Police Service (SAPS) will appear in court on charges of defeating the ends of justice after they were recently arrested for manipulating crime statistics at their police station in Tzaneen in the Limpopo Province. An internal police investigation into anomalies in this station’s crime data allegedly found evidence that these commanders had deliberately tried to improve the crime statistics for their precinct by recording murders as ‘inquests’ and downgrading cases of housebreakings to the less serious crime of trespassing. This was apparently done to make it look as if the police station was effectively tackling crime in the area. Given the misuse of the crime statistics as general measures of police performance by the SAPS, manipulation of this data at the police station level has emerged over the years as a particular problem.

This is one of the reasons that along with their annual release by the Minister of Police, many people ask, “How accurate are the crime statistics?” This is an understandable question when members of the public continue to experience and hear about heinous deeds, but are told that the statistics say crime is going down. Indeed, no crime statistics anywhere in the world are considered to be scientifically reliable measures of a crime challenge. Policy makers and analysts require additional independent sources of information to obtain a fuller picture of the crime situation. It is therefore very good news that the South African government has made the decision to conduct regular annual national Victims of Crime Surveys (VOCS) to be undertaken by Statistics SA (Stats SA​).

Stats SA highlights that VOC survey intends to meet the following three key objectives:

To explore the nature, extent and patterns of crime in South Africa, from the victim’s viewpoint.
To identify victim risk so as to inform the development of crime prevention and public education programmes.
To obtain citizens’ perceptions of services provided by the police and the courts so as to contribute to improvements in the work of the criminal justice system.

Fortunately, South Africa already has a precedent for this type of survey. Stats SA undertook the first of such surveys in 1998 and the Institute Security Studies (ISS) subsequently conducted VOCS in both 2003 and 2007. The 2011 Victims of Crime Survey therefore provide the firsts set of independent national level statistics in four years from which to assess the recent police crime statistics. It must be noted upfront however, that VOCS data cannot be directly compared with police statistics as they are collected, collated and analysed differently. The current VOCS data is based on a large nationally representative sample of 29 754 households. It therefore can provide useful insights into national and provincial trends and perceptions. However, the data cannot be used to obtain an understanding of crime challenges occurring at a local community level. Police data on the other hand is collected locally (i.e. at each police station for a local precinct), and then aggregated upwards to provide provincial and national statistics). Therefore, VOCS statistics are primarily useful in comparing SAPS statistics at national, provincial levels and potentially large metropolitan areas.

Importantly, the 2011 VOCS survey findings support the trends in the SAPS statistics that show a decrease in South African crime levels. For example, the 2007 VOC survey found that 57% of households thought that crime had increased in their areas over the previous four years while only 18% thought that crime had decreased. In the 2011 VOC survey, the picture looks substantially different. Only 37% of households think that crime has increased in their areas compared to a larger proportion of households (42%), who believe the level of crime has decreased. Therefore in 2011 more people thought that crime is going down when compared to those who think that it is increasing. This shift is also reflected by perceptions of safety. In 2007, only 22% of survey respondents felt safe walking alone in their areas after dark. This has increased notably to 37% of respondents in 2011. An additional useful component of VOCS data is that it provides breakdowns of the data geographically and demographically. Therefore we know that Gauteng residents most likely perceive crime as having decreased in their area with 52% of households sharing this view. We also know that females in the Free State province feel the least safe walking alone in their areas after dark compared to all other provinces. Only 6% of females in the Free State feeling safe walking in their areas alone after dark compared with the 24% of females who feel safe in the Limpopo Province.

The survey also finds that most households (53%), thought that residential burglary or housebreaking was the most prevalent crime in their areas, followed by home robbery (50%), street robbery (41%) pick-pocketing (29%) and then assault (21%.) When asked who was most likely to commit crime in their area, 61% of households stated that local people from the area committed it, while 32% believed that the perpetrators were South Africans from other areas. Only 7% of the respondents thought that the perpetrators of crime in their areas came from outside South Africa. This challenges the oft-stated sentiment that South Africans think that foreigners are responsible for the crime in the country. There were quite large provincial variations in perceptions in relation to this. A vast majority of Northern Cape households (82%) said that crimes were committed by locals living in the area compared with only 43% of households in Gauteng who thought so.

When asked why perpetrators commit property crime, the largest proportion of households (58%) thought that it was because of “real need” due to poverty or unemployment, rather than greed (46%). Only one in five respondents (20%), thought that other factors such as drug use was the motivating factor.

Given this sentiment, which is consistent with the findings of previous VOC surveys, South Africans generally don’t believe that the solutions to crime lie with the criminal justice system. When asked what government should spend money on to reduce crime, to two thirds (66%) of households thought that it should be on social or economic development. This compares starkly to the much smaller 21% who thought that more money should be spent on policing. When it comes to personal security, half of the households took some type of “physical measures” to protect their homes, 11% relied on private security, 9% belonged to a “self help group” and only 5% took to carrying a weapon for protection.

The survey revealed interesting trends in relation to corruption committed by government officials. Unsurprisingly as it is in line with previous surveys, the biggest culprits were traffic officials. Of those who were asked for a bribe in the previous year, more than half (53%) came from traffic officers to avoid a fine. Worryingly, this type of corruption has increased since 2007 when substantially fewer (33%) respondents reported this type of corruption. Traffic related bribery was most prevalent in Gauteng (62%), the Western Cape (57%) and then followed by the Eastern Cape (56%). Since 2007, corruption was found to have increased in relation to the following public services: policing, obtaining housing, a drivers licences, water or electricity, court related services and medical care. On the positive side, corruption was found to have decreased in relation to obtaining identity documents or passports, pensions or welfare grants, prison visits, customs, schooling and employment.
The VOCS data provides a rich set of information from which to assess trends and dynamics in citizens’ experiences and perceptions of crime and the criminal justice system. However, the statistics in their current form only tell us what the attitudes and experiences of citizens are, not why. For example, while we know that women in the Free State feel more unsafe than in other provinces, it is now necessary for research to start focusing on telling us why this is the case. Similarly, why corruption is increasing in relation to traffic offences, policing and housing. Only once we understand why these changes have occurred, will policy makers be able to know what can best be done about it. Fortunately, Statistics SA is also willing to make the full data sets available to whoever wants to analyse them. This adherence to the important constitutional principle of transparency will enable larger numbers of pubic, private and civil society organisations to better understand the nature and dynamics of crime and thereby be better able to contribute towards solutions. Currently, we spend over R83 billion on the criminal justice system and approximately R40 billion on private security each year in reaction to crime. Very little funding is available for research into understanding crime and how to prevent it. Until this is corrected we will continue to spend money on the consequences of a high crime rate without doing what is needed to prevent it from happening in the first place.

Written by Gareth Newham, Programme Head, Crime and Justice Programme, ISS Pretoria Office