Published:  2014-12-30 Views:  1477
Author:  karelnel
Published in:  Security/Sekuriteit
Crime Stats For Harties Up To 2013 - Part 9 : Measurement

Crime Stats for Harties up to 2013 – Part 9: Measurement

The reader should, after reading previous articles in this series, understand that the sheer volume of crime statistical data makes it difficult to comprehend. Even if it is interpreted by mathematical and statistical techniques, it is still difficult to draw conclusions. Consider for example the simple question: Did the incidence of crime increase the past ten years? It is difficult to answer because some crimes may have increased while others may perhaps have decreased. There are also other complicating factors like, for example, a change in population size.

There is a truism that says: “To measure is to know”. If we do not measure properly, we shall only have perceptions about a matter, and different people will have different perceptions about the (same) matter. Enters the question: How can one measure crime? This article will deal with this question by addressing two possible instruments: A crime score card and a crime monitoring sheet.

 Crime Score Card:

How can one devise a single metric that will characterize the ESSENCE of the crime picture in an area in a single figure? Such a figure, if consistently applied over the years can then give a good view of the crime impact trend as a whole.

Obviously such a metric need not include all crime statistics but should be representative of all the most important crime types. The previous article has shown how a selection of the most important of the high impact crimes can be made, based on the Pareto distribution. The previous article also showed that the “Impacts” of these crimes can be calculated by the product of its incidence- and severity figures. To even make the metric even more representative, different crime types can be lumped together and a combined Severity index calculated. See table 1 in the example below.

Without going into the mathematics, the example below shows an application of this method, based on the 2013 figures and the assumed crime severity indices.

Example: Crime score card compilation:

Table 1: Highest Impact crime groupings selected and their severity indices.



Crime Types Included

Relative Impact (Assumption)



Murder, Attempted Murder and Culpable Homicide




Assault GBH & Common Assault& Sexual Crimes




Robbery with aggravating circumstances, Common robbery, Burglary Residential, Burglary Non-residential, Robbery residential and Robbery Non-residential




Theft of Motor vehicles, Theft out of Motor vehicles, Other theft, Shoplifting




Drug related & Driving under the influence







Applying this grouping and severity figures to the data of the last 10 years produced the following crime score cards.


Figure 1a: Hartbeespoortdam                             

Fig 1b: North West Province                     


Fig 1c:  Republic of South Africa

From the above (Keep in mind the underlying assumptions) it can be seen that there was a significant downward crime impact trend in Harties since 2009 and the SAPS is to be credited for a job well done. A positive contributing factor may also be the active involvement of the community in the CSS.

Crime Monitoring Sheet:

It is not sufficient to only have a global metric that will measure crime impact consistently on an annual basis. The crime statistical data must be put to work on a continuous basis in the fight against crime. An instrument, used in Quality Engineering, called “Statistical Process Control” can be used in conjunction with crime statistics for this purpose. In Statistical Process control, control charts or control sheets are compiled that will show at the earliest possible opportunity when a manufacturing process is beginning to deviate from its intended outcome, by making use of samples.

Although future crime incidences cannot be predicted, where clear trends are seen in the data one can deduce that processes are embedded in the community that drives these types of crimes. Such trends may therefore be indications of what to expect. It is even possible that realistic “targets” could be set to be achieved for the reduction in crime, based on the historical data.

The historical database can be used as a benchmark to measure future statistics to determine when significant shifts do occur. Starting off from the historical database, data should be refined and limits more accurately determined to afford better discrimination when significant changes in crime patterns do occur. This should be done on a MONTHLY basis to discover significant trends AT THE EARLIEST POSSIBLE OPPORTUNITY. (Obviously also taking the seasonal variation into consideration)

A simple Excel-based template can be developed to support the public in recording and interpreting and displaying the crime information in a user-friendly way. This will make the information more accessible for crime management purposes. Figure 2 gives an example of such a reporting format. This format can automatically indicate when significant increases in crime types are experienced.

 Figure 2: Example of a Crime Monitoring Sheet

A way forward?

This series of articles was compiled by strictly looking at the statistical parameters of the published public domain data. Although interesting trends can be discerned, all these are in the domain of uncertainty and should not be seen as predictions, but rather as indications that can be helpful in the understanding of possible future scenarios. No inputs from any policing or security practitioners were solicited during the compilation of these articles. The value of the articles can therefore be greatly enhanced if these graphs and trends can be workshopped with experienced partners from the SAPS.

In an interview during 2014 Lt Col van der Walt indicated that the SAPS have a sophisticated crime management system at their disposal that can give them all the above capabilities and more, and that support of the public in this respect is not needed. The public, however, has a right to sensible crime information, and if the SAPS are not in a position to supply such information to the public, the community must collect and interpret information for themselves.

The following may be considered for the future:

  • A crime analysis work group should be established

  • The uncertainties in the data and crime categories are clarified

  • The figures should be carefully recalculated and conclusions are to be more accurately drawn
  • Sensible metrics and management tools must be compiled as is described in this article

  • That one or two benchmarking studies are also done (comparable towns) and other supporting statistics (eg. Population figures) are also found to make wider benchmarking possible.

  • The results are presented to the SAPS for possible scrutiny and improvement

  • The results are presented to the community to inform them about the historical and current state of affairs – possibly a Harties “Crime Fighting Conference”?

  • That guidelines for further analysis of crime data be researched and implemented where possible

  • That the information is made available on a website for use by the community.

Closing Remarks:

This concludes this series of articles on the 2013 crime statistics. Future articles will deal with additional topics as well as perhaps the significance of the 2014 published crime statistics.




Directors: S.A.M Smit & J.F Koster - Copyright ©2019 HartiesNet (Pty) Ltd - our Disclaimer