Teasing out crime trends from crime statistics is often bewildering, with some crime types going up and others down. The delayed releases by the South African Police Service (SAPS) also mean that what the nation is currently experiencing is not always what is reflected in the data.
We are now in the part of the year where crime tends to rise, so it feels as if “things are getting worse”. But the signal coming from the data to September was that we may have turned the corner on violent crime.
In an environment of low and variable trust in policing, which in turn affects reporting rates, which in turn makes the trends in crimes such as sexual offences difficult to interpret, murder is a strong signal to look at.
This is because an increase in reported sexual offences could mean survivors are reporting more — perhaps because of better response via interventions such as Thuthuzela Care Centres. Or it could mean that there are, in fact, more sexual offences. It is difficult to discern which interpretation is more at play.
Decrease in murder
By contrast, murder is still generally well-recorded in South Africa and is thus the best indicator of trends in violent crime. The reduction in murder associated with the first year of the government of national unity (GNU) has been large and unequivocal, and takes effect in the very first quarter of data after the sixth administration left office.
This was sustained for the first full year (July 2024 to June 2025), and indeed accelerated in the quarter to September. What drove this trend, and will it continue?
A clue can be found in the trends for the years immediately before the formation of the GNU. After the Covid-19 lockdown ended, murder accelerated, and once load-shedding took hold in earnest, this trend was amplified. However, around this time it became noticeable in the data that the Western Cape began to diverge in quarterly trends from the rest of the country. Before, it had closely mirrored the national seasonal trends, albeit at a higher level.
What are the candidate reasons for the diversion? There are obvious consequences of load-shedding that might worsen crime. This includes crime effects arising from, among other things, street lights and security devices not working, and restricted deployment due to lack of electricity.
Load-shedding and murder rate
Though load-shedding affects the whole country, Cape Town was able to protect its customers from one or two stages via the Steenbras hydroelectric scheme. The city supplies electricity to about 70% of its residents, which in turn comprise 66% of the provincial population.
Analysis in mid-2024 revealed that in the Western Cape, there has been no relationship between the national number of days of load-shedding and the murder rate. However, for the rest of the country, the number of load-shedding days per quarter was statistically significantly associated with the number of murders per quarter, with 32% of the variation in murder predicted by the number of load-shedding days, and every additional load-shedding day in a quarter being associated with 20 extra murders.
Load-shedding ended abruptly in May 2024. In the first quarter of the GNU, which was also the first quarter of no load-shedding, the number of murders dropped by 1,151, an incredible 17% compared with the same quarter the previous year. But in the same quarter the previous year there had been 74 days of load-shedding. The absence of those 74 days could accordingly account for much of the change.
Some lack of load-shedding effects are longer term, so further reductions can be expected. Indeed, the entire first year of the GNU recorded lower murders, with the total reduction a staggering 3,145, an 11% decrease in the murder rate.
The reduction was also present in the Western Cape, but to a far more muted extent (5% down), as the province would have been less affected by the lack of load-shedding. It was even less pronounced in Gauteng (2%), large tracts of which continued to suffer significant load-reduction.
The Western Cape and City of Cape Town would also be keen to point to their joint contemporaneous “Leap” intervention. The Law Enforcement Advancement Plan involves city-employed law enforcement officers deployed with the SAPS in areas with high murder rates. This may have contributed to the load-shedding and post-lockdown effects being less pronounced in the city than in the rest of the country, and thus the prior divergence.
Indeed, the murder signal gives hints of a slight effect (albeit with still gravely concerning trends) in Leap areas, compared with areas that don’t benefit from the intervention. However, in the Western Cape a spate of internecine gang violence beginning in August 2025 has made it the only large province to clock an increase in murders in July to September, relative to the same period in 2024.
The continued failure to allocate sufficient SAPS members to areas of high murder rates in that province — despite a court order to reassess allocations — remains infuriating.
Will the reduction in murders continue? A happy coincidence of the GNU and the reduced number of murders invites the drawing of a causal link. However, if there is a link between reduced crime rates and governance, it is likely to stem from another (associated) dimension.
Virtuous cycle
What we learnt during the 2000s is that reductions in crime result in a virtuous cycle, in which less violence leads to less violence still, in a relatively rapid fashion: in 10 years the murder rate steadily decreased until it had halved by 2010.
But we know what happened next. The governing party was essentially taken over by a ruinous faction that bled the state dry, corrupting the police and justice system and ushering in murderous load-shedding.
There has been some justice in that. Since 1994, crime and the fortunes of the governing party at the polls have been closely linked: an increase in murder compared with the year of the previous election has been associated with a decrease in voting share, and vice versa — until the governing party no longer achieved a majority of the votes cast in the 2024 general election.
If the country can maintain the momentum it has now built up over five quarters, it may well manifest in a virtuous reduction in the violence cycle that benefits all — including, perhaps, the parties participating in the GNU, at the polls. This will require that the most venal are kept away from the reins of power — they are still circling.
• Dr Redpath is a senior researcher at the Dullah Omar Institute.
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.