Back to Africa Check

ANALYSIS: Why crime stats trends don’t recommend Cele (or anyone) for senior police leadership

Former SA national police commissioner Bheki Cele’s recent appointment as police minister has been cause for much debate.

Fuelling discussions are his controversial statements on officers’ right to “shoot to kill”, his role in the re-militarisation of police rank structures and his involvement in improper conduct and maladministration around the lease of police headquarters.

Some have sought to defend his appointment on the grounds that his period as police commissioner, from 2009 to 2012, saw a decrease in the recorded rates of certain crimes.

News site News24 quoted Dr Johan Burger, formerly from the Institute for Security Studies, as saying “the stats show that, during his three years as police commissioner, the most serious crimes either declined or stabilised. Ever since he left, crime stats have been on the up.”

Online publication Daily Maverick claimed that the Cele period “was characterised by significant average annual decreases in nine more policeable crimes”. It concluded: “So there is proof of the pudding in the eating. Something happened during the Cele period which proves that policing in this period had significant reduction results.”

But there are a number of reasons why these arguments should not convince us.

Stats tell a more complex story


It is easy to pick just those data points that support a certain argument. In fact, the trends in different crimes over the past twenty or so years have been very different.

Shown here are some graphs of the long-term recorded trends in the national rates of various crimes, indicating the broad periods in which different national commissioners were in office.





Going by these charts, it seems that Cele’s leadership period was associated with no particular deviations from longer-term recorded trends in residential burglary and vehicle theft, two non-violent property crimes.

Similar patterns are seen in a number of other crimes, including assault with intent to cause grievous bodily harm and common assault. These began declining long before and have continued declining after the Cele years.

For these crimes, there is no evidence that anything particularly unusual happened during the Cele period or of significant changes since his departure.

On the other hand, these years do seem to have marked the beginning of turning points in the trends in murder and aggravated robbery.





The recorded rates of certain sub-categories of aggravated robbery did significantly decrease in especially the first half of the Cele period, followed by increases since. As these graphs show, this is particularly clear in rates of house robbery and carjacking.





Correlation not causation


Even if Cele’s leadership period were associated with exceptional, unambiguous decreases in all crime types, it would not follow that the one thing was responsible for the other.

There are countless, sometimes humorous, examples of spurious correlations. This happens when two or more events or variables may seem to be causally related to each other, but their relationship is actually due to coincidence or the presence of unseen other factors.

A number of other things happened in the years directly after former police commissioner’s Jackie Selebi’s removal from the post in early 2008. Xenophobic violence broke out in various parts of the country, the elite crime-fighting unit the Scorpions was disbanded, the country hosted the Confederations and FIFA World cups, it endured a significant recession, and Richard Mdluli was appointed head of police crime intelligence and later suspended following criminal charges.

The years since Cele’s removal from office have seen the Marikana massacre, the unfolding story of “state capture”, the coming of age of the “born free” generation, significant political shifts in a number of major metropoles and so on.

Any one or combination of these, or many other events, could as easily be held responsible for observed trends in crime. Research has suggested that crime rates can be influenced by factors including police legitimacy, political stability, state legitimacy, demographics, inter-generational shifts in norms, and labour market conditions.  

Cele’s leadership may well have had nothing to do with it.

Recipe for deception & poor service delivery


Another reason to hesitate in accepting the argument that Cele should be credited with crime declines is that the evidence is decidedly mixed that most traditional policing activities are effective at preventing crime before it happens.

Although police are not powerless over the short- to medium-term, their role in determining the long-term ballpark and sustained trends is likely limited. Crime rates often fluctuate with no relationship to changes in policing.

On the other hand, something that is well within the police’s power is to discourage victims from opening official dockets, to “downgrade” reported offences to the least serious crime type possible and to lose or destroy dockets.

Holding individual police officers or the institution as a whole responsible for reducing the number of crimes recorded is a recipe for deception and poor service delivery. Victim and community satisfaction and case outcomes are better measures of whether the police are doing their jobs.

Pinning crime rates to certain police leaders contributes to this disincentive for victim-centred, professional and high-quality policing. It makes our police worse at their jobs. We should stop doing it.

Anine Kriegler is a researcher and doctoral candidate in criminology at the University of Cape Town. She is a co-author of A Citizen’s Guide to Crime Trends in South Africa.

Add new comment

Restricted HTML

  • Allowed HTML tags: <a href hreflang> <em> <strong> <cite> <blockquote cite> <code> <ul type> <ol start type> <li> <dl> <dt> <dd> <h2 id> <h3 id> <h4 id> <h5 id> <h6 id>
  • Lines and paragraphs break automatically.
  • Web page addresses and email addresses turn into links automatically.
limit: 600 characters