The SAPS crime statistics – measuring what exactly?

It’s time for the Great Annual Bad Maths and Funny Logic Debate — you know, the one we have every year in which the SAPS tells us what a great job they’re doing combating crime and we all pick holes in their arguments in response. At one level there’s need of this, at another level it’s a thoroughly sterile debate that gets in the way of deeper questions such as what exactly are the crime statistics (not) measuring? Who bears what responsibility for preventing violence? And by what methods is prevention accomplished? We could proliferate these questions endlessly but let’s start with questions of measurement.

At this point all the crime statistics tell us is how many people are willing to invite the criminal justice system into their lives. They are not any sort of gold standard indicator of how much crime is actually taking place. Take rape for example. According to the SAPS the rate of reported sexual offences declined by 45.3% in Gauteng between 2003/04 and 2012/13. Now before we congratulate the police on a terrific job well done consider the research undertaken in Gauteng in 2010 by Gender Links and the Medical Research Council. This found that one quarter of women in the study had been raped in the course of their lifetimes, while almost one in 12 women had been raped in 2009. But only one in 13 women raped by a non-partner reported the matter, while only one in 25 of women raped by their partners reported this to the police. In other words, what the Gauteng police statistics are measuring is an alarming silence on the part of women about the sexual violence they experience.

We could also treat the rape statistics as a proxy indicator of women’s confidence in the criminal justice system. To draw on research again, this time a case study provided by Tshwaranang Legal Advocacy Centre of one rural locality in Mpumalanga. This analysis of 252 rapes reported between 2005 to 2007 found that more than half were reported by girls under the age of 18. This finding runs counter to both national and international trends showing the majority of rapes to be reported by adult women. Either the researchers stumbled upon a world exception, or produced findings skewed by adult women’s non-reporting. Once you take into account the fate of those adult women’s reports in comparison to those of children, then the latter explanation seems more likely. First, the police made fewer arrests in adults’ cases and also spent less time investigating their complaints. Second, once at court prosecutors withdrew women’s matters more often and sooner than children’s cases and prepared fewer cases for trial. As a result only one of the 120 rapes reported by adult women during this three year period resulted in a conviction.

In a small rural community where many people know each other’s business, these sorts of outcomes provide no incentive to report.

Domestic violence is no less difficult to measure from police statistics. To begin with there is no category of crime labelled “domestic violence”. Instead such violence is captured within a range of different crime categories including rape, murder and assault with intent to cause grievous bodily harm. And because the police don’t capture the relationship between the perpetrator and victim in their statistics, we have no idea how much domestic violence is being reported to the police, or in what form. And just like sexual crimes, domestic violence is significantly under-reported. Back to the example of Gauteng: Between April 2008 and March 2009, a total of 12 093 women (or 0.3% of the adult female population in the province) reported an assault by an intimate partner to the police. By contrast, during the same time period 18.1% of women in the province reported an experience of violence at the hands of intimate male partners to researchers. As with sexual crimes, only a fraction of domestic violence is coming to the police’s attention; a decline in numbers is not to be celebrated.

With these limitations in mind, organisations belonging to the Shukumisa Campaign have called for a meeting with the SAPS to discuss collating the sort of statistics that would be useful in planning for services and programmes. First up is age-disaggregated data. There’s been a recent outcry about the rape of elderly women for example — is this on the increase? And where are these rapes occurring? This is important for purposes of research and intervention. In addition, how many incidents of sexual violence perpetrated against lesbians are being reported to the police? And how many men and boys are reporting experiences of rape?

Additionally, there are 59 different types of crimes that can be classified as sexual offences. The police need to start providing figures for each of these different offences and not one figure overall. This will help us to see how many of the new crimes introduced by the 2007 Sexual Offences Act are being used in practice (especially those committed against children and people with disabilities), as well as to what extent the police are still charging people under the remnants of the 1957 Sexual Offences Act (which now chiefly deals with crimes related to sex work, such as keeping a brothel).

The questions then revolve not only around how the police calculate their crime figures but also around what those figures mean and imply for policy and practice. Let’s debate the statistics that count.