Prime Times
Analysing When Crime Occurs
A detailed analysis of available data on when crime occurs suggests that monthly crime totals vary considerably from area to area. The same cannot be said when crime is plotted according to day of the week, where there is remarkable consistency. In contrast, the time of day when crime occurs varies both according to the area and crime type. These fluctuations in crime raise more questions than can be answered with the available information but are important in the formulation of policing responses.

Victim surveys conducted by the Institute for Security Studies in Pretoria and Durban asked respondents to name the month. day and time when they were victimised. Cape Town data was only available in relation to the time of day when crime was committed. This data - which has remained largely unanalysed - suggests some surprising conclusions. Among other uses (although much more work is required), the data may hold important findings for the design of police shift patterns and the timing of particular crime prevention interventions.

Few useful comparisons can be drawn between monthly peaks in crime across the cities. Thus, for example, car theft has been prominent in Durban in June - 15% of all incidents being recorded in that month over the last five years. The same does not apply to Pretoria where only 8% of all incidents of car theft occurred in that month. Similar conclusions can be drawn in regard to comparisons of all crime types between the two cities. This suggests that monthly variations in crime are likely to be area specific, for example, on seasonal fluctuations or conditions, - perhaps the larger number of tourists in Durban at specific times of the year - rather than on any nationally determined set of factors.

Figure 1 Which day of the week violent crimes were committed in Durban and Pretoria
Source: ISS Victim Survey

There is however a remarkable degree of consistency across Pretoria and Durban in regard to the day of the week on which crime occurs. Given their consistency these figures have been combined and represented in Figures 1-3. In particular, violent crime - most notably murder and assault - are far more likely to occur on Fridays and Saturdays than at any other time of the week.

Figure 2. Which day of the week property crime happened in Durban and Pretoria
Source: ISS Victim Survey

While property crimes are more likely to show a more general dispersal across days of the week, there is still a prominent concentration of crime on Fridays and Saturdays. Property crime, which entail the use of violence - such as hijacking and mugging/robbery - are also generally consistent with these patterns.

 

Figure 3. Which day of the week violent property crimes happened in Durban and Pretoria
Source: ISS Victim Survey

There are a number of possible reasons for this phenomenon. Violent crime - although it is difficult to generalise - often occurs around places of entertainment where alcohol is available. These places are usually frequented on weekends. It is more difficult to explain why property crime also concentrates on weekends, given that the opportunities for such crimes are relatively similar, if not greater - for example, the numbers of cars on the road in the case of hijacking - during the week. It is possible however that these patterns reflect not only opportunities for the commission of offences but also availability of offenders. This explanation of course may be equally unsatisfactory given that it is not easy to identify when offenders are at work and thus unavailable to commit crime.

There is some consistency as to when crime occurs during the day across the three cities. Thus mugging/robbery is most likely to take place between midday and 6pm when opportunity - for instance, the number of people walking on city streets - is highest (Figure 4).

Figure 4. Mugging / Robbery - Time of day.
Source: ISS Victim Survey

Incidents of assault are also relatively consistent when measured in this way: the majority of cases in all three cities take place between midday and midnight. Such similarities are much more difficult to establish is respect of a number of other crime types. Both burglary (Figure 5) and hijacking (Figure 6) show few if any parallels between the cities. Burglary for instance, appears more likely to occur in the morning in Cape Town and at night in Pretoria and Durban. 

Figure 5. Burglary - Time of day
Source: ISS Victim Survey

Hijacking shows a more distinct variance: the majority of hijackings in Durban appear to occur in the mornings from (6am to midday), in Cape town in the afternoons (midday to 6pm) and in Pretoria at night (6pm to midnight).

Figure 6. Hijacking - Time of day
Source: ISS Victim Survey

Why these different patterns apply across the cities is difficult to determine from current data.  What is clear however is that while monthly and hourly patterns suggest a degree of variance which may relate to the specific conditions in any city, weekly crime patterns begin to point to important similarities in crime patterns across cities, deserving of more focused.