SAPS clarifies the use of ratios in analysis of Crime Statistics

Pretoria - The South African Police Service has noted some of the media reports and questions from various media about the use of ratios and raw numbers when reporting crime statistics.

Although some of the media houses, such as the Daily Maverick, have demonstrated an understanding of this issue, SAPS thought it prudent to issue the statement clarifying the matter.

Both the ratios and raw numbers (actual case numbers) were provided to the media and loaded onto our website, as it has always been the case. The use of ratios is not intended to deliberately confuse matters or hide anything. The SAPS uses ratios as it a common international practice when doing comparisons.

When using ratios, it is possible to find that although a certain crime category may have decreased, it may show increase when considering the raw figures. There is nothing untoward.

The key issue is that since crime is a social phenomenon, one has a take a long time view as well as factor a key variable of the country's population in order to improve analysis. If one does not factor population change, it will result in an incorrect analysis and improper understanding of crime statistics. Normally, one would expect that as the population grows so should crime due to the increased opportunities, particularly contact crime. However, longitudinally, this has not been the case with South Africa, at least in the past nine years.

Example for calculating ratios: Total sexual offences

During the 2011/2012 financial year, 64 514 sexual offences were recorded in South Africa and the population size was 50 586 757 in that same period, the crime ratio istherefore calculated as follows:

Crime Ratio or rate = (Crime figure/Population size) X 100 000

 = (64 514/50 586 757) X 100 000

 = 0.001275314 X 100 000

 = 127.5

The figure tells us that during the 2011/2012 financial year, 127.5 persons among each 100 000 people were victims of sexual offences.

 During the following financial year (i.e. April 2012 to March 2013) 66 387 sexual offences were recorded and the population size of South Africa was 52 274 945. The crime ratio is calculated as follows:

Crime Ratio or rate = (Crime figure/Population size) X 100 000

 = (66 387/52 274 945) X 100 000

 = 0.00126996 X 100 000

 = 127.0

The latter tells us that during the 2012/2013 financial year, 127.0 people out of 100 000 were victims of sexual offences.

The two figures above (127.5 and 127.0) are not percentages, but crime ratios.

If one wants to determine the net movement or change in crime that is whether a real increase or decrease in crime occurred, crime ratios are best suited for the purpose. Crime ratios or rates equalise population growth and therefore allow comparison of real incidence of crime over time. It is for this reason that crime ratios are used.

Percentage Increase or Decrease

Percentage increases or decreases indicate the direction of crime. Crime can move in one of two directions or remain at about the same level. In order to determine whether crime levels have gone up or down, one needs the crime figures recorded during two corresponding time periods. The two periods are usually the period currently under review and the corresponding period in the past. 

Using the sexual offences ratios, the percentage increase or decrease recorded during the period under review (April 2012 to March 2013) is calculated as follows:

= (Current Period - Previous Period) X 100

 Previous Period

= (127.0 - 127.5) X 100

 127.5

= -0.5 X 100

 127.5

= -0.0039 X 100

= -0.4%

So, the outcome, when applying ratios is that a year-on-year comparison shows that there was a -0.4 reduction in sexual offences. This does not change the fact that year-on-year, when looking at raw figures; there is an increase in sexual offences.

As it relates to the use of population figures, SAPS calculates crime ratios using the population figures provided by Statistics South Africa (StatsSA). Only StatsSA is mandated by legislation to collect analyse and disseminate demographic information, specifically population figures. 

StatsSA publish mid-year population estimates every year. The mid-year population estimates are based on the previous census results. All the mid-year population estimates from 2002 to 2011 were based on the 2001 census. The SAPS used the 2011 mid-year population estimates provided by StatsSA to calculate the crime ratios for 2011/2012. This is consistent with what we have been doing all along and it allows us to compare apples with apples when analysing the data.

Statement issued by SAPS Head: Corporate Communication, Lieutenant General SM Makgale, September 20 2013

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