NEWS & ANALYSIS

2013/14 crime stats: How is the SAPS really doing?

Former CIAC head Chris de Kock writes that in measuring police performance one must focus on police-able crime categories

THE 2013/2014 CRIME STATISTICS: WHAT DO THEY REALLY SAY ABOUT SAPS' PERFORMANCE?

This analysis clearly indicate that SAPS:

Failed to provide security to the people of South Africa in their homes and on the streets (public/street robbery shows the largest increase of all crimes without any mentioning on the website).

That murder, of which a large proportion is in fact more police-able, had its two most significant increases since the establishment of SAPS in the last two financial years and this put S.A. backward to 2009/2010. Internationally murder is used as an indicator of stability due to its data integrity.

That in fact the only reduction impact that the SAPS had on more police-able crime was on stock-theft and on theft of motor vehicle.        

1) INTRODUCTION.

From August 1995 to April 2013 the author of this document was responsible as the head of the Crime Information Analysis Centre (CIAC) of the SAPS (South African Police Service) for inter alia the gathering, analysis and dissemination of crime statistics. In this capacity he and the team working under his command was part and parcel of all the public releases of crime statistics by the Ministers of Safety and Security and later Police and the National Commissioners in that period.

The only exception to this in the period that he was working at SAPS was the 2011/2012 release of crime statistics in September 2012 which was done under the direction of the Chair of the National Commissioner's task team on crime statistics then Gauteng Provincial Commissioner, Lieutenant- General Mzwandile Petros.

Dr. De Kock retired at the age of 60 from the SAPS in April 2013, but as a crime and violence researcher and analyst and Sociologist he will only retire when his health does not allow him to continue. His participation in the debate on the wrong ratios that SAPS used in the 2012/2013 crime statistics, which they refused to change on the official SAPS website up to now and the substantive input (report and evidence) that he made to the: "Commission of Inquiry into Allegations of Police Inefficiency and a Breakdown in Relations between SAPS and the Community in Khayelitsha" are only two examples of the work after retirement.

Since his retirement he is actually in a much better position to contribute to a more objective understanding of crime and violence in South Africa, since he does not have to fulfil all the administrative and operational functions that he had at SAPS and as an independent analyst he does not serve any master, except maybe the normal subjectivities of a white male, middle class South African, which he can also neutralize after more than 40 years of been a sociologist and researcher.

2) ORIENTATION.

The purpose of this short, first analysis of the 2013/2014 crime statistics is to cut through the overload of statistics and arguments and counter arguments which accompany it in the first hours after the release of crime statistics. It is to come to the essence/heart of the crime statistics and to at least conclude if the South African Police are making a difference to the crime situation in the country.

When crime statistics are released, at least once a year, over the past two decades, since the establishment of the SAPS, there were always claims and counterclaims from the police themselves and those that analyse the statistics that the police play an extremely small to no role in crime reduction or that it is solely responsible for reduction.

Usually the SAPS would claim victory when the statistics look good and when it does not look good claim that it is not a primary role player in crime reduction. Those who can't wait to criticize SAPS will do the opposite, when crime look good it is the people of S.A. and when it look bad it is all the fault of SAPS. This debate does not bring clarity.

This short analysis will be done from the following departure points:

This analyst moves from the departure point that although all crimes are police-able, some are more police-able and others less police-able. Policing basically include three groups of actions namely: a) intelligence actions, b) prevention actions and c) detection actions.

If all of these actions in all three groups can be used to reduce a crime category, it is a more police-able crime. If only detection and those intelligence actions which support detection can be used to address a crime category it is reactive policing and a less police-able crime category.

Where crime occurs in the private/non-public sphere (like in the bedroom of a couple's house or in a club/shebeen/bar or shop or in cyberspace) and police prevention can basically not occur it will usually be less police-able crimes.

For the purpose of this first analysis the following less police-able crimes are not included since a police service on its own cannot really reduce most of them through intelligence and prevention:

-Assault GBH (90%+occur in private spaces, between people knowing each    other).

-Common assault (the same as assault GBH).

-Arson and malicious damage to property which mostly go with assaults.

-Sexual offences (the same as with the assaults, although maybe 75 %+ know each other. Sexual offences are a mixed bag of more than 50 different offences of which some, as for example those related to sex work, are actually police detected crimes and can very easily be "managed downward").

- Theft: other (there are small proportions like cable theft which is more police-able, but the majority of cases occur in private spaces-e.g. a ring taken from a bedside table or a cellphone lifted off an office desk).

- Shoplifting (perpetrators "arrested" by security and customers and then handed over to police).

- Commercial crime (the police can't patrol cyberspace where most of these cases occur these days or even the paper files of a private company).

- Crimes detected as a result of police action where an increase actually show that the police were doing a good job. These three crimes will not be include here, but it will be necessary to write something about them at a later stage since there is more to it than just police input produces an output.

The intention of this analysis is not to minimize the seriousness of crimes like sexual offences, assault GBH, and commercial crime. It is just saying that all of these crimes are less police-able. Because large proportions of these crimes can only be addressed after-the-fact by police detection, and not intelligence based prevention, police performance can't be easily measured against them.

This analyst will never minimize the horror of a gang rape or a baby rape which in all probability will end in murder and even mutilation of the body. Some gang rapes (like hate rape or so-called corrective rape) and serial baby rapes, can be prevented through intelligence work, but in the crime statistics as released by the SAPS all of these are mixed up with boyfriend-girlfriend type rape, which is far harder to prevent, and sex work where the police may turn a blind eye, because if they open cases it will increase the number of reported sexual offences.

Murder and attempted murder, because of their heterogeneous nature are included in the present analysis. Shortly after the release of the 2013/2014 crime statistics some analysts referred to the fact that 80% of murders and attempted murders are social in nature, but these findings are quite outdated.

In the 2008/2009 SAPS Annual report on page 10 and 11 the findings of a docket analysis done on a representative sample of 1 348 murder dockets of the period 1 April 2007 to 31 March 2008 are provided and according to that only 65, 4% were social in nature and the rest (34, 6%) encompassed a wide range of different motives, such as murder as a result of: another crime (usually robbery), reaction to crime (e.g. legal self-defence, in the line of duty and vigilantism), intergroup conflict ( e.g. gang, taxi, faction/clan fighting), and hate crimes (e.g. xenophobia, homophobia).

All of these crimes are more police-able and this was a good 5 years back. In the meantime there are many indications at station level that this 35% proportion may have grown. Even from the docket analysis which was done in the late nineties it was clear that of attempted murders at least 45% were due to non-social motives like robbery, group conflict, vigilantism, and hate crime.

Murder is included because of the importance of these crimes and the fact that the more police-able (non-social) element seems to be growing and if the police had taken the necessary intelligence and prevention steps it will in all probability show a reduction in the total murder and attempted murder categories.

Furthermore as a sociologist, this analyst can say that it is improbable that the social conditions and factors that generate the social proportions of murder and attempted murder had suddenly increased over the past two years, especially if it is considered that assault (both GBH and common) actually decreased.

In this analysis only categories and sub categories of crime will be used. For example the different kinds of aggravated robberies are used and not the category aggravated robbery as such. No groupings of crime like TRIO, contact, and serious crime will be used.

When it comes to analysis the whole idea is to disaggregate phenomena to its smallest parts since that is the only way to come to a better understanding of the phenomena. Unfortunately when we do analysis outside of the police or even inside on a national and provincial level we can't break the legally based categories further down, without doing docket analysis, as the station analyst can do on a daily basis.

One of the greatest challenges to proper crime analysis in South Africa is that the legally based categories of crime are very wide and heterogeneous, but on the other hand it should be remembered that the pre-2000 crime code list with detailed crime codes also had a strong downside, namely invalid and unreliable crime statistics.

Broad crime groups (categories grouped together) are often used to hide increases of critical more police-able categories amongst large less police-able categories. A decrease in contact crime as a result of decreases in the large proportions of assaults may hide major increases in smaller categories of more police-able crimes like aggravated robbery, murder and attempted murder.

The time periods reflected in the table of comparisons and its discussion was chosen as such, since it makes historical sense. The analysis cover a period of a decade from 2004/2005 to 2013/2014(with a baseline of 2003/2004).

It is actually not wise to cover a period before that since the moratorium on crime statistics were from July 2000 to May 2001 and a decision was taken then to only use the crime statistics after the 2001/2002 financial year which was actually 2002/2003 while the baseline here is 2003/2004.

The other historical reason why this decade is used, is that it covers two MTEF (Medium Term Economic Framework) periods. Which is basically two administrations, namely the second and last administration of President Mbeki and the first administration of President Zuma. To refer to two MTEF periods will not make sense to the reader and that is why the reference will be to the two administrations.

During the second Mbeki-era the SAPS was managed by National Commissioner Jackie Selebi for the largest part and policing was based on the National Crime Combating Strategy (NCCS) with its strong emphasis on intelligence-led policing at station level, and constant monitoring at all levels of the impact on crime reduction.

During the first three years of the first Zuma administration the police for the largest part was managed by National Commissioner Bheki Cele. Since right from the onset of this period it was preparation for World Cup 2010 and the event itself a strong emphasis was put on crime reduction through focused police visibility at station level and constant monitoring.

Despite promises of at least keeping the crime levels at the levels achieved during this event it seems that the same levels of focussed police visibility could not be sustained and shortly after this General Cele left.

It should also be emphasized that there may have been other factors like lower unemployment and a groundswell of patriotism which increased downward pressure on crime.

In the last two years of the first Zuma administration the National Commissioner was General MV (Riah) Phiyega, with the exception of the first quarter of 2012/2013 financial year. This period right from the start was characterize by events like Marikana and other prolonged violent strikes and public violence, incidents of alleged police brutality, and a range of political scandals which often compromised the police. It was also characterize by many changes in both national and provincial management.

This analyst would actually prefer to use ratios in the calculations of trends. However since the SAPS so far, even with the new crime statistics on the website, refuse to update the 2004/2005 to 2011/2012 ratios with the 2011 census and new midyear estimates. In the process they have built in a 2,2% benefit nationally and in some provinces as much as a 9,0% benefit. As a result he will now have to use total figures. This actually poses no problem since if figures are used consistently they are scientifically sound. It is just that they reflect marginally more negatively on the state of crime in the country.

In the sixth column of the table below, under the heading "Achieved/failed to achieve ...",  it is indicated if the crime reduction targets, as set out in the "Annual Performance Plan 2013/2014" of SAPS, were achieved or not and if not achieved how far in percentage points they were missed.

From pages 1 and 27 of the Performance Plan it can be deduced that contact crime had to be reduced by 6, 8%, TRIO crime by 6, 2% and serious crime by between 0, 8% to 1, 4%. Although this analyst differs both with the use of crime statistics to measure police performance, as it may corrupt the figures, and with the categories used to measure it, it nonetheless remains an agreement with the people of South Africa.

3) FIRST ANALYSIS OF CRIME STATISTICS AS RELEASED ON 2014/09/19.

The percentage increases/decreases in raw statistics/frequencies for 2013/2014 of all more police-able crimes are compared to different time periods in the attached table. Please note that the percentages in columns 3 to 5 are average annual decreases/increases for the specific period. If this table is scrutinised the following deductions per more police-able crime can be made:

Public/street robbery which was provided on the police website in previous years but for some or other reason is not available on the 2013/2014 website. This needs an explanation by the SAPS, especially since this is the single largest increase of all the categories and sub-categories of more police-able crime in the table.

The increase is a significant one of 14, 2% for 2013/2014 following the increase of 4, 2% in 2012/2013. It should be noted that in the ten year period reflected in the table there were only three years in which this crime increased and two of those were the largest increases - which were the 14, 2% increase in 2013/2014 and the 4, 2% in 2012/2013 respectively.

This is disturbing for this is the single largest proportion of aggravated robbery and specifically the kind of robbery which affected the working classes of the mega townships and informal settlements like Khayelitsha, Nyanga, Tembisa, and New Brighton and the old Central Business Districts like Johannesburg, Pretoria and Durban.

These are the people who are robbed at knife and gun point of their meagre salaries on their way home from work on public transport. Sometimes they pay the price with their life. In all probability this lies at the root of vigilante reaction as in Khayelitsha.

It is also very significant that in both the Mbeki administration period as well as in the first three years of the Zuma period (when SAPS was mostly under management of General Cele) this crime decreased by 6,3% and 6,6% respectively.

In the 2004/2005 to 2008/2009 period TRIO crimes (carjacking, house robbery and business robbery) escalated with an average annual increase of 15, 3% while public/street robbery decreased with the annual average of 6,3%. The explanation for this is that as SAPS took control of the streets and public spaces the criminals were pushed into targeting residences, businesses and vehicles.

During the first three years of the Zuma administration inroads were made on all the different subcategories of aggravated robbery except business robberies. Now, in the last two years, it seems that SAPS again lost control both of the streets and all the other subcategories.

The last column of the table indicates that SAPS missed the performance agreement target with a dismal 21 percentage points. But in all probability as in the previous years this may not even be discussed in-depth anywhere.

The TRIO robberies of house robbery, business robbery and carjacking increased by 7, 4%, 13, 7%, and 12, 3% respectively. These increases were much higher than the respective increases of 7, 1% 2, 7% and 5, 4% in 2012/2013 for house robbery, business robbery and carjacking.

It is also significantly higher than the decreases and increases for the respective crimes during the first three years of the Zuma period, when house robbery decreased with an average annual decrease of 3, 0%, carjacking decreased with an  average annual  decrease of 12, 2%, and business robbery increased with an a average annual increase of 4, 9%.

The only comfort, when it comes to the three TRIO crimes, which the management of SAPS can find in the table, is the average annual increases of house and business robbery during the Mbeki era, but then it should be remembered that the baseline were much lower in that era and that the largest proportion of aggravated robbery (street/public robbery) were decreasing consistently and significantly.

Last but not least is that business robbery, carjacking and house robbery failed the performance targets by 20 percentage points, 19 percentage points and 14 percentage points respectively. In any private business this would have required serious explanation.

Truck hijacking, which is a highly organised/syndicated crime increased by 5, 1% during 2013/2014.This at least is a smaller increase than the 14, 9% during the previous year (2012/2013). And over the last two years (the NATCOM average) it is in line with the increases in the Mbeki era, although during the first three years of the Zuma era an annual average decrease of 14,3% was obtained.

From past experience it was established that truck hijacking, CIT-robbery and bank robbery fluctuate according to the rise and fall of syndicates. If a new syndicate appears the trend may increase quite dramatically and if they are arrested or disappear for another reason the trends may decrease very significantly.

The sudden jump of bank robbery from 7 to 21 cases in 2013/2014, after years of decline from 144 in 2007/2008 to 7 in 2012/2013 also illustrates the point and is in all probability due to a new or old syndicate or two which has appeared or reappeared.

Common robbery basically stabilized on a marginal increase of 0, 6%. During the last two years(NATCOM average of 0,8%) it basically did not increase or decrease, while in the Mbeki and the first three years of the Zuma administration it decreased by 7,6% and 3,5% respectively.

The increases of 5, 0% in murder and 4, 6% in attempted murder, when taken together with the increases in every subcategory of aggravated robbery, are the biggest shock of the 2013/2014 crime statistics.

The reader should again be reminded that murder is accepted internationally as the most reliable crime trend because to put it bluntly it boils down to a body count. There can virtually be no underreporting.

Murder, with the exception of an increase of 3,5% during the 2006/2007 financial year, which was characterized by high levels of violence during the security guard strike, constantly decreased during the ten year period which is analysed in the table.

During the Mbeki administration it decreased with an annual average of 1, 7%, while during the first three years of the Zuma administration it decreased with a very significant 4, 7% annually. The 4,2% and 5,0% increases in murder over the past two years are actually the two largest increases in the decade under review here and actually since the creation of SAPS in 1994.

These increases over the last two years rewind South Africa back to 2009/2010 (the Confederation Cup year) in which there was a 7,2% decrease in murder  and the 2010/2011 World Cup year in which there was a 5,3% decrease in murder.

Likewise attempted murder decreased with 4, 9% and 11, 0% in respectively the Confederation Cup and World Cup years and the increases of the last two years also rewind S.A. back to 2009/2010. In fact over the past two years S.A. should have moved forward to levels of 13 700 and 12 900 murders and attempted murders respectively, if the promises of both the previous Minister of Police and National Commissioner had been realized, namely that a new target/bar was set in 2010/2011 and that we should not move backward.

It must again be emphasized to the reader that murder rates/figures are accepted internationally as reliable and are often used as an indicator of stability in countries. On this crime category South Africa has to move forward and not backward!  

In an hour or two after the release of the crime statistics crime analysts and experts provided some hasty explanations for the increases in murder and attempted murder with which this old analyst does not agree and which can be misleading.

This analyst and his team at CIAC (Crime Information Analysis Centre) of the South African Police Service, over a period of 15 years, did quite a bit of docket analysis, which was mostly published in quarterly CIAC reports and SAPS annual reports.

Originally (back in the late nineties and early two thousands) 80% plus of all murder dockets were indeed of people who knew each other (friends, lovers, husbands and wives, neighbours, etc.). The crime, which was usually committed under the influence of alcohol, generally started as an argument about liquor, money, sex, and or family matters, which then developed into a assault ending up as a death or deaths (murder).

A retrospective cohort docket analysis of dockets registered in 2000 also revealed that, over the following five year period, in 34,6% and 5,7% of cases of completed prosecutions a murder conviction was not obtained but rather a culpable homicide or serious assault one (see SAPS Annual Report 2009/2010,p.9)

So in the late nineties and early 2000's this analyst would have agreed with the explanations of his colleagues that murder is a less police-able social phenomena and that the only thing that the police can do is to investigate the dockets, find the culprits and see to it that the perpetrators of murder get their day in court.

But then slowly but surely in later docket analysis it changed to only 65,0% of dockets which are of this nature (see section above). And that was also 5 years back. Now five years down the line it may very well be that 40% to 50% of all murder dockets are those which are the result of other crime, mostly robbery aggravated, intergroup conflict (taxi, gangs, factions etc.), reaction to crime ( self-defence and vigilantism) and hate crimes( homophobia, xenophobia). And one thing is for sure, these can and should all be prevented by the police. So it is actually seriously misleading to exempt the SAPS on the basis of outdated docket analysis.

The question which should also be asked is did police management at all levels, but more specifically at National and Provincial Commissioners level, not sit down after at least the annual figures for 2012/2013 were finalised towards the end of April 2013 and do a proper introspection of what lead to the 2012/2013 increases of respectively 4, 2% and 10, 1%?

They should have actually identified and discussed these increasing trends long before April 2013 - even as early as July 2012 - but at least in preparation for the 2012/2013 crime statistics release. One would have thought that they would put their heads together and called in the analysts and be prepared for the media.

All of this could have prevented the second (2013/2014) shock of significant increases in aggravated robberies, murder and attempted murder.

In a report and verbal evidence before the Khayelitsha Commission this analyst stated that after analysing minutes of the three stations that make up Greater Khayelitsha (Khayelitsha Site B, Harare and Lingelethu-West) he can only come to a conclusion that policing in that area is not intelligence-led but policing by chance.

He also concluded that the concept of intelligence led policing can fail because:

1) that because of a lack of skills, experience and or motivation, the CIAC/CIO at station level don't deliver a Crime Threat Analysis product which clearly indicates to the station commander and operational management the what, where, when, why, and who of each crime category in that station;

2) the station's operational management cannot operationalise the information/intelligence that the CIAC delivers because of a lack in skills, experience, wrong attitude, and/or motivation; and

3) the station management receive very good quality information/intelligence and do operationalization, but have very limited or no resources to execute the plans. If everything happens according to the book and the specific crime is still stubborn, as the National Commissioner put it so fittingly, then it requires going back to the drawing board, checking the information/intelligence and re-operationalise.

After this release the question should be asked loud and clear, and which the SAPS should answer, is whether the problem of policing by chance, which was identified in Greater Khayelitsha only a problem there or also in Nyanga, Inanda, "station Eastern Cape" and "station RSA" to just mention a few examples.

Because if that is the case we can again accept that no introspection of these 2013/2014 crime statistics was done at least in April 2014 and that next September (2015) we may be in for another rude shock.

This also bring this analyst to a last remark on the statement made by, for example, the ISS that the main culprit for the increase in all kinds of aggravated robbery, and I then assume all the non-social murders and attempted murders, is the so-called Richard Mduli saga.

About this so-called saga this analyst, which was there when it started back in 2011, did not say anything in the past and will not say anything in future, except where it is legally required. But although it possibly may had or have a certain effect on higher order national and provincial intelligence collection it is highly unlikely that it had any negative influence on the CIAC/CIO structures at station level which also fall under the command and control of the station commander, where the bulk of crime analysis is supposed to be done and the collection which occurs at cluster level and which is supposed to be based on the CTA which is produced by the CIAC.

All of these local and cluster activities, which actually forms the basis of intelligence led policing, experience problems which are experienced worldwide and over many years. So it was there before the so- called Mduli saga and although this saga should have ended long ago - before it did all the damage on a higher level - the disappearance of the saga will not on its own improve intelligence led policing at the lowest level where it can make a difference.

Burglary: residential with a marginal decrease of 0, 6% in the reporting year strictly speaking just stabilised after a 6, 8% increase in the previous year. The same applies to burglary: non-residential with its 0, 0% and 5, 1% increase in 2013/2014 and 2012/2013 respectively. It should again be noted that in the related crimes of business and house robbery there were significant increases.

The only indication that SAPS had any possible reduction impact on more police-able crimes are the decreases in stock-theft and theft of motor vehicle over the past two years. Stock- theft decreased respectively with 6,2% and 3,4% in 2013/2014 and 2012/2013, while motor vehicle theft decreased with 1,2% in 2012/2013 and 2,6% in 2013/2014.

Dr. Chris De Kock is a Consultant/Analyst: crime, violence and crowd behaviour.

Click here to sign up to receive our free daily headline email newsletter