Crime Statistics methodology: SAPS, Stats SA & independent crime analysts briefings

Meeting report

The Chairperson noted that the Committee would not have a meeting on 6 June 2017 due to the all-day sitting of the House.

Mr Z Mbhele (DA) requested that the Minister or Deputy Minister of Police should appear before the Committee next week to brief them about the Independent Police Investigative Directorate (IPID) and South African Police Service (SAPS) matter which seemed to be escalating in terms of litigation.

The Chairperson noted the request. The Committee would have to make a formal request to either the Minister or the Deputy Minister as this was indeed an important matter. The Committee noted that there were eight police officers in KwaZulu-Natal (KZN) on charges of corruption, defeating the ends of justice and forgery. It was good to see SAPS taking a great stance in getting rid of the “rotten apples” and deal firmly with corruption. The Committee also noted the visit by the President and Minister of Police to Elsies River, Cape Town and the handing over of the satellite police station. There were allegations from Elsies River community members that police members in the area were in cohoots with local gangsters. The Committee believed that this matter should be investigated by IPID in the Western Cape and by the police to ensure that the police are able to regain the confidence of the community.

The violence and killing of women is also a matter of great concern and it was good to hear that the matter will be debated in Parliament this week. There is progress in the Stellenbosch case and the police were able to apprehend all the suspects. This showed once again the good work done by the detectives. We also saw good results in the Khayelitsha Regional Court in dealing with the R20 million abalone syndicate and the suspects were handed a combined sentence of 127 years. This will send a strong message that organised crime does not pay. The Committee last week raised VIP Protection in Bloemfontein and requested a report from SAPS within seven days. The Committee had not received any feedback on the matter although seven days had already passed. It was totally unacceptable for those accounting to the Committee not to furnish the Committee with written responses.

Mr Mbhele requested the Committee Secretary do a follow-up on other written responses that still needed to be forwarded to the Committee as this was about ensuring accountability.

Mr L Ramatlakane (ANC) also appreciated that SAPS was willing to root-out corrupt police and “rotten apples”. It was concerning to continually hear about the problem of gangesterism in the Western Cape and this was something that needed to be dealt with vigorously.

SAPS Crime Statistics - Methodology for Compiling Crime Statistics: SAPS briefing
Major General Norman Sekhukhune, SAPS Head: Crime and Research Statistics, indicated that crime statistics focus on reported crime perpetrated within the borders of South Africa at the 1 144 police stations, including satellite stations and ports of entry. The statistics would comprise of crime reported by the victim, witness, third parties and detected by SAPS during any policing activity. The administrative data collection process involves recording the criminal incident. The source document used for this purpose is a case docket. The Community Service Centre (CSC) Commander peruses the case docket for correctness and completeness before it is registered on the Crime Administration System (CAS) or Integrated Case Docket Management System (ICDMS). The first statement of crime must contain four elements of crime:
• Principle of legality, incident constitutes an offence
• A human being had to have performed the act
• Unlawfulness, the conduct contravenes a statutory requirement or a common law rule
• Culpability, i.e. intention or negligence.

Maj Gen Sekhukhune said that all crimes reported are recorded and this was irrespective of when the crimes were committed. The elements of crime are assessed to ensure that the correct classification or charge is allocated to the incident. The number of counts associated with the identified offence are then determined. The preliminary investigation and inspection of the incident are conducted to ensure that the information reported is accurate and complete. Crime is categorised together in group crimes with similar characteristics and this classifies the nature of crime and facilitates the understanding of the various crime categories. There are five broad categories of crime: contact crimes, contact related crimes, property related crime, other serious crime and crime dependent on police action for detection. All the crimes recorded at the stations within a particular cluster should collectively add up to the total of the crimes recorded in that cluster. The crimes recorded in the clusters within a province, should collectively add up to the total crimes in that province. The crime recorded in the nine provinces, should collectively add up to the national level.

Maj Gen Sekhukhune pointed out there had been a number of quality improvement initiatives:
• Crime data quality management to enhance the crime statistics quality at station level, the Crime Information Managers (CIMAC) conducts daily quality checks and any discrepancies are discussed and rectified at the Station Crime Combating Forum (SCCF)
• The National Crime Registrar performs the same function of quality assurance
• The Technical Working Team (TWT) was established to be responsible for quality assessments of the annual crime statistics prior to release, implementation of actions to address quality concerns and recommendations
• The previous two annual Crime Statistics reports were quality assessed and the process endorsed by the Statistician General

In conclusion, the Crime Registrar is developing policies and standards for the whole statistical value chain of crime statistics production. The approved policies include policy on crime statistics and fundamentals of crime statistics analysis. The draft annexures to the policies include classification of crime for statistical purposes, crime counting rules and standard guidelines on crime registrar station products and Standard Operating Procedure for using them.

Independent Crime Analyst on Crime Statistics compilation
Dr Chris De Kock, an independent crime analyst, said he had been involved in three committees of inquiry into the generation, analysis and dissemination of crime statistics. He noted the difficulties about classification for a particular crime and the crime categories used. There is a huge dilemma around aggravated robbery, sexual offences especially rape, and murder such as so-called farm murders, femicide/gender based murders, homophobic and xenophobic murders. There are also counting rule dilemmas. There is significant under-reporting of crime. This can all lead to a trend analysis nightmare. He suggested this universal phenomenon of under-reporting of crime should just be accepted and it should be emphasised that the crime statistics are based on reported and registered crime. At least for the past years, one has the excellent National Victims of Crime Surveys (VOCS) of Statistics South Africa.

Dr De Kock stated that under/non-registration of crime (there can also be over-registration) can be either unintentional or intentional. There are numerous cases where the police officer convinces the victim that there was no crime; or the police officer accepts that there was a crime, but only “registers” it in the incident book/register; or the police officer opens a docket, but does not list all crimes and/or counts of crimes on docket cover; or the police officer does not enter all crimes and or counts on the CAS (Crime Administration System). Statistics are drawn from CAS for a specific crime, station, and for a specific period. The station statistics add up to cluster level, cluster level to provincial level and provincial level to national level. He noted the importance of the drawing of the crime statistics at the same intervals.

He gave an analysis of the SAPS approach to crime statistics in South Africa since 1994. It was clear that government wanted reports that can scientifically explain the crime phenomena. However, from 2011/12 to 2014/15 crime reports used any trick in the book to convince the public that SAPS is achieving its reduction targets. South Africa lost most of the gains achieved during the period 2004/05 to 2011/12. The 2015/16 crime report was the first acceptable report since 2010/11.

The essence of analysis is that the analyst break down the crime to it smallest parts and then put it together again to understand the phenomena. Each subcategory of a legal category should in all probability be addressed differently with different strategies for different kinds of murder, rape and aggravated robbery. 
For the past four to five years SAPS did the opposite. They group more- and less policeable crimes together in groups of crime, such as contact crime, contact-related crime. SAPS use these broad crime types which in the case of contact crime include both less and more policeable crime to hide increases in more policeable crime. With this they are fighting the crime statistics and not the crime. Analysis is not to “hide”, but to come to an understanding of the what, where, when, how, why and who of each crime that should be prevented and detected (if not prevented). 

It is high time that the more serious, more policeable, crime categories like murder, rape and aggravated robbery are broken down to operationally useful subcategories and that it is registered as such. Without this, problem-orientated policing is not possible and intelligence-led policing is not fully possible.

He gave a number of suggestions on how crime statistics could improve and these included:
• The validity of crime statistics can be improved by members of the Crime Registrar reconciling the crime registers with dockets and checking if all the crimes and their counts were listed on the docket and then registered on CAS. This should be done daily. This was a recommendation of the Petros Committee.
• These station members should fall under the command and control of the Provincial Crime Registrars who fall under the command and control of the National Crime Registrar who reports directly to the National Commissioner.
 
Statistics South Africa briefing on collaboration with SAPS
Mr Harry Thema, Executive Manager: South African National Statistics System (SANSS), Stats SA, indicated that the crime statistics committee commissioned a report in 1998 that highlighted areas that needed improvement and specific attention and these included improving SAPS capacity to disseminate crime information. Interventions were put in place for system improvement and these focused on improving CAS and data retrieval, a new operational system for statistics plus improved CAS for docket tracking. A 2009 crime statistics concept note stated areas of concern, and Stats SA’s proposal to improve crime statistics in the country was to prioritise the quality and capacity gap first and the information gap can follow. SAPS approached the Statistician-General in May 2011 to assist them on the road toward crime statistics eventually being accorded the status of official statistics. In August 2011, a framework of engagement was adopted by the Stats SA / SAPS task team as the de facto approach towards the collaboration.

Mr Thema pointed out that there were number of issues that were brought to the fore by the National Task Team in 2011 and these included:
• No approved policy for the management of crime information
• Counting rules for the SAPS
• Margin of error in the recording of crime

SAPS and Stats SA signed a Memorandum of Understanding (MoU) in 2015 and the introduction of the clearing process by the Statistician-General. In terms of the way forward, the International Classification of Crime for Statistical Purposes (ICCS) has been endorsed by the United Nations Statistical Commission in March 2015 and the Commission of Crime Prevention and Criminal Justice in May 2015. The adoption of the ICCS at the national level will require the attentive translation of offences as defined by the legislation into ICCS categories, with careful consideration of the full act/event descriptions and explanatory notes. The adoption of this international classification will facilitate data integration throughout the various chains (i.e. SAPS and Correctional Services). This would also contribute to the promotion of data sharing as well as enable statistical data to be comparable.

Institute for Security Studies briefing
Mr Gareth Newham, Institute for Security Studies; stated that crime statistics could not be the scientific representation of crime levels and this was the case everywhere in the world and not just in South Africa. Likewise crime statistics could not be used to measure the effectiveness of the police on the ground. It is essential for crime statistics to be released more regularly but this should not be incorporated as a measure of crime levels. Communities should know crime levels in their areas so as to have policies in place to deal with crime at a local level. There should also be a map that provides communities with hotspots for crime and ways to deal with the problem. There should be a cultural change at station level where station commanders are willing to accurately report crime and those stations that are reporting crimes more frequently should not be punished.

Discussion
The Chairperson welcomed the presentations and asked about the skills level within Stats SA as this was not mentioned in the presentation. The Committee should also be briefed about the role of the station commanders in reporting and compiling crime statistics. It would be important to know if the recent malicious crimes committed against young women would be reported and the case dockets would be open. What is the total number of people conducting crime research within SAPS? It was stated that reported crimes are discussed by the station crime combating forum on a daily basis. An important question would be about the measurements that are in place to ensure that this is happening at every station in the country. The Committee should also be briefed about the role of the station commander in supervising that this is actually taking place.

Mr Z Mbhele (DA) asked SAPS for clarity on the four elements of crime mentioned in the docket as some elements looked similar. SAPS should provide the Committee with disaggregated data about crime that would provide a clear and accurate picture of crime in the country. It would be important to know if there is a tracking and recording of the crimes committed against the LGTBI community. There should be localised and customised crime statistics available on a weekly basis for the community to be able to analyse and become aware of trends of crime. The SAPS presentation was unclear in tracing the movement of a case from the district to the court. SAPS should utilise Geographical Information System (GIS) to improve crime detection.

Ms M Molebatsi (ANC) asked about data integrity and the process involved in ensuring that all reported crimes were recorded. It was unclear what would happen if the reported crime did not fall under the four elements of crime. The Committee should be briefed on whether there were enough police per district to be able to deal effectively with crime in the country. What are the strategies in place to address the problem of unreported crimes?

Ms M Mmola (ANC) asked if SAPS was ready to release 2016/17 crime statistics. It was unclear if crime statistics had any impact on SAPS planning in dealing with crime incidents around the country.

Mr L Ramatlakane (ANC) commended SAPS for collaborating with various stakeholders like Stats SA in ensuring that there was data integrity in the crime statistics. The Committee should be briefed about capacity within SAPS at various stations. What could be the remedy that SAPS could incorporate to have perfect categories or subcategories of crime? The reality is that there are many categories of crime and it would be difficult to narrow down the categories of crime. There are indeed many cases of unreported crime in the country and it would be difficult for SAPS to solve this. ISS should perhaps explain how SAPS could incorporate unreported crime as the crime statistics were based on reported crime.

Mr M Mncwango (IFP) asked about the capacity in place to provide support to encourage the reporting of crime. What happened when prosecution of a reported crime ended up being unsuccessful in court?

Mr J Maake (ANC) asked for clarity on data capturing at station level and the training level that was required. What was the method being used for capturing crime data? SAPS indicated that in some countries crime is also recorded after adjudication had been completed while in other countries it was the other way around. What are the advantages and disadvantages of capturing reported crime before the completion of the adjudication process? Who was responsible for the collection of crime data? It would unfair to blame the police for unreported crime. What should police do in this context? It was indeed correct that there should be crime statistics at local level. Who are the people responsible for data capturing at the local level?

Ms L Mabija (ANC) said some reported crimes are regarded as “unfounded” by the police although they had taken place and this was something that needed to be addressed. The fact that some crimes are categorised as “unfounded” was discouraging communities from reporting crime.

Lt Gen Gary Kruser, Deputy National Commissioner, SAPS, responded that reported crimes are discussed at station level and this was an important matter that was being taken into consideration. The reported crimes are also analysed and perused at cluster level and there was more capacity at this level.

The Chairperson interjected and asked if there were any possible ways in which the Deputy National Commissioner would know if a particular station had a meeting and what the agenda discussed was.

Lt Gen Kruser replied that the systems in place did not allow him to have such information and this was a practice that SAPS was currently building together with Maj Gen Sekhukhune. The coordination at the station level would provide SAPS with a better picture of what was happening at every station in the country. SAPS was busy with a centralised approach where a reported crime at the station level would go directly to the National Commissioner together with Strategic Management. SAPS is required to go back to the person who reported a crime and explain why a certain crime did not fall under the four elements of crime. SAPS was working with an SMS system where an SMS would be sent to the person who reported crime explaining if a certain crime did not fall under the four elements of crime. This is a way of providing quick feedback on the reported crime.

Brigadier Mmakosha Mahlangu, Section Head: Crime Statistics and Research, SAPS, replied that the Crime Administration System was for the capturing of reported crime and assigning the case number. The ICDMS was an improvement of CAS as it is able to project all the case dockets in the system and this was something that was not happening in CAS. The ICDMS is a fully developed system but it was being implemented in phases and that was one of the reasons SAPS was using both systems at the moment.

Lt Gen Kruser added that SAPS was using both of the systems because they had not been implemented in all police stations in the country. The CAS system would be immediately phased out as soon as ICDMS is implemented at all the stations.

Maj Gen Sekhukhune explained that SAPS was using crime statistics based on the legally defined crime but the legally defined crimes were not sufficient for SAPS to be able to understand the crime patterns that were happening in different localities. The subcategories of crime were meant to make crime statistics more comprehensible to the people. In essence, the introduction of new categories and subcategories is meant to enhance the understanding of crime statistics. All crimes that are reported to the police in South Africa are registered and then after the registration there is an investigation as to whether a particular crime did indeed take place. There are indeed cases where a reported crime would be unfounded and this usually happened in cases where a person would suspect theft without exploring other possibilities. There are now common cases where men will lie and say they had been hijacked while in reality they did not sleep at home. These cases are then classified as unfounded.

Maj Gen Sekhukhune mentioned that every crime that is brought to the attention of the police is registered and this was a South African way of doing things. In other countries such as United Kingdom, a reported crime is only registered after the investigation. Trust in the statistics in places like the United Kingdom is at a more advanced stage than in South Africa. It is preferable for South Africa to keep the current system where crime is registered first and subsequently investigated.

Mr Newham clarified that the statement that crime statistics is not the scientific representation of the crime levels was not meant to be a controversial statement and certainly not a criticism of the police as this was global phenomenon. There is Stats SA that could help the country in improving the reliability of crime statistics by conducting surveys on household crime. SAPS should find ways to encourage people to report crime so that it could be reflected in the statistics and not punish police stations that are bedevilled by crime. SAPS should also look at the categories and subcategories that are more policeable. Communities are aware that some levels of crime are happening because of multifaceted problems like unemployment, poverty, drug and alcohol use. It would be helpful for the Minister of Police to invite other ministers like Minister of Social Development and Justice and Constitutional Development to find ways to collaborate in fighting crime. There should be a major shift in thinking and we must stop regarding police as solely responsible for crime prevention.

Ms Moola asked for the process followed if the captured crime was given the wrong case number.

Lt Gen Kruser responded that the system in place automatically generates the case number and there is rarely a case where a captured crime can be given a wrong number. There is an SMS that is sent to the person who reported a crime and this SMS contained the case number.

Ms Mabija asked if there is a system in place to detect cases where a person who reported a crime was not provided with a case number.

Lt Gen Kruser responded that the station commander would be the one to deal with cases where a person had reported crime and was not provided with a case number. The system was the one generating the case number and not the individual.

Ms Moola explained she had personally witnessed someone who was given a wrong case number and this was a matter of concern.

Lt Gen Kruser said he did not have evidence as to whether the person who captured the crime on the system did not want to report the crime or there was an actual mistake. The station commander is the person to refer to in those cases and there is also complaints line where people could voice their grievances.

Mr Thema replied that the police ratio per district could not be determined by Stats SA as this differed from place to place. Police visibility was dependent on the propensity of crime in the particular area. There are recommendations put in place from time to time depending on the threat analysis. Stats SA needed to be as independent as possible and this was something to take into consideration. The internal side of SAPS is not the responsibility of Stats SA but the main focus is on improving the methodology of collecting crime statistics. It is also true that crime statistics from the side of the police will always be a contested issue but it is better to have the crime statistics than not having them at all. The trends of crime are able to assist the nation to lift up those issues that required improvement. There are a lot of methodological improvements that are being witnessed in how SAPS is collecting reported crime and improving data integrity.

Mr Thema said that the discussion that the country should have is whether crime statistics should continue to be handled by SAPS or an independent body. The integrity of the figures is dependent on the independence of the person conducting the statistics. Even if Stats SA were to decide to take over the collection and release of crime statistics, it would still have to rely on SAPS for information as they are the one with the administrative records. Although there is a lot of improvement by SAPS, as indicated before there are also disappointments. Stats SA would continue working with SAPS on improving data integrity.

Lt Gen Kruser replied that there is no tracking of hate crimes at the moment but SAPS could do a docket analysis to come up with data on this. SAPS was doing analysis of crime and looking at reasons why some crimes are committed more regularly. SAPS was conducting further research as to why certain communities do not have trust in the police to be able to address this problem. There is research at the moment where SAPS is looking at customer satisfaction and this was based on the survey that was released. Based on the survey and research, SAPS was looking at ways to enhance customer satisfaction.

Maj Gen Leon Rabie, SAPS Head: Strategic Management, replied that SAPS was working on ways to bring police stations closer to the people as a way of improving the reporting of crime.

Mr Risenga Maluleke, Stats SA Deputy Director-General: Statistical Collections and Outreach, indicated that the problem of geographical location in South Africa was a contested issue. The country would need to find ways to deal with geographical alignment and for the police to be able locate places where there are no street numbers as in rural areas. Stats SA was addressing this issue of geographical alignment. Stats SA is able to pick up crimes that were never reported at police stations as Stats SA would personally speak with the victims themselves. However, the biggest challenge facing SAPS and Stats SA is on the reporting of social crimes, crimes occurring within the same family or the same neighbourhood.

Lt Gen Kruser replied that SAPS was using GIS for crime detection and the focus now was on using Google Earth on top of the GIS and this was giving people better operational capacity.

Maj Gen Sekhukhune clarified that the matter of data integrity started at station level and this was something that was taken care of by the station commander. There are various stages that the reported crime would go through from being reported at the station to being captured on the CAS system. There is also a crime registrar sitting at cluster level that undertakes functions of perusing the case docket and verifying recorded information. There is quality checking at every stage from the provincial and national level and this was essentially meant to improve data integrity. The quality checking at national level is usually undertaken during the quarterly release of the crime statistics.

Lt Gen Kruser said that SAPS was indeed doing fieldwork and this would usually be undertaken at crime scenes and SAPS was working together with Maj Gen Sekhukhune in this terrain. SAPS would do the analysis of patterns of crimes committed in a particular area and this is to determine the motive for the crime and the people that are involved in crime.

Dr De Kock explained that the problem of underreporting was a universal phenomenon and found all over the world. There is not much that could be done to deal with the problem of underreporting of crime except the monitoring and surveying of victims of crime as is being done by Stats SA. SAPS needs to look at the reasons for underreporting of crime and one of those reasons is that nothing was being done about the reported crimes. Vigilantism was common in Khayelitsha and this was because of this lack of trust in the police. The problem of vigilantism had increased in the country from virtually zero a few years ago to about 10% of the murders that take place. SAPS should include other subcategories of crime to improve on reporting and prevent vigilantism. There should be subcategories for murder and rape and this is to implement strategies of dealing with these crimes based on the correct stipulated category. Crime statistics is the very basis for crime intelligence.

The meeting was adjourned.