Crime statistics belong to the public, not the police

Having Stats SA validate and release crime figures would separate the statistics from police performance and depoliticise the data.

Published on 22 October 2025 in ISS Today

Anine Kriegler
Senior Researcher, Justice and Violence Prevention, ISS Pretoria

The release of South Africa’s quarterly crime statistics has again been postponed. The quarter one 2024/25 figures were initially scheduled for dissemination at the end of August, then 16 October, and are now delayed with no new date confirmed.

This underlines the need for a more stable, transparent system. Crime data is a public good. It should be routinely released to support cross-government decision making, enable evidence-based crime prevention, and help communities respond effectively.

South Africa’s violent crime crisis is not just a policing problem, but a shared responsibility, requiring coordinated input from many sectors of government and society. This fact is widely recognised by the government, the police and analysts. Yet every quarter, it’s the National Commissioner of the South African Police Service (SAPS) who reports the latest trends to Parliament. This is not a productive way to interrogate the data or frame the problem.

The current arrangement forces police leadership to account for complex socioeconomic trends that extend beyond their mandate. It also undermines public trust. Because crime statistics are used to measure police performance, both internally and externally, the police face strong incentives to under-record incidents. This creates a conflict of interest that can distort the data, discourage timely sharing and undermine services to victims.

There is no ‘correct’ model for how crime statistics can or should be made public. Countries handle this differently depending on their institutional capacities, legal frameworks and policy priorities. In many places, arrangements have evolved as systems mature and expectations for transparency grow.

The current system forces police to account for socioeconomic trends that extend beyond their mandate

For example, in 2016 the SAPS transitioned from presenting crime statistics annually to quarterly. The Institute for Security Studies advocated for, and welcomed, this change. Reliable crime data is essential for effective policy. It should be regularly released by a neutral, technically competent body that is free of operational pressures.

One option is to transfer the responsibility to Statistics South Africa (Stats SA). This isn’t the only possible model, but it illustrates how a more stable, transparent system could look.

The SAPS and Stats SA have a memorandum of understanding to enhance the quality and integrity of crime statistics. It commits the two institutions to cooperation on standardising classifications, improving verification and quality assurance, and building capacity to ensure more reliable, consistent data.

The SAPS has already invested considerable effort in strengthening its data systems and analytical capacity. This proposal builds on that foundation. The next step could be to formally designate Stats SA as the official body for validating and releasing crime statistics at the station, provincial and national levels.

The model has precedent. In Canada, crime statistics are produced and released by Statistics Canada. In Australia, they are published by the Australian Bureau of Statistics. In the United Kingdom, responsibility moved from the Home Office to the Office for National Statistics in 2012 to address concerns about politicisation of the process.

In all these countries, official statistics are complemented by victimisation surveys. Because the surveys capture incidents not reported to the police, they offer a more complete picture of crime harm. Stats SA already has this capacity through its annual Governance, Public Safety and Justice Survey and its broader constitutional mandate.

Using crime statistics as a performance measure incentivises the police to under-record incidents

The official release of crime statistics could follow a model similar to other national datasets. Stats SA would publish the data quarterly without fanfare, accompanied by clear methodology and basic trend analysis. Once a year, Stats SA could lead a more structured publication and interrogation cycle, combining the year’s statistics with victimisation survey results and in-depth analysis of long-term trends and drivers.

Parliament could convene dedicated portfolio committee hearings, drawing on the SAPS for operational insights, as well as departments such as social development, health, justice, and basic education. This creates the proper forum to interrogate underlying causes, policy performance and cross-government responsibilities.

This approach wouldn’t fix the limitations of police crime statistics. The figures would still reflect only what is reported to and recorded by the police. Many incidents never reach the police because people may lack trust in the justice system, fear retaliation or consider the matter too minor to report. This happens worldwide, which is why official statistics almost always understate the true scale of crime.

However, consistently pairing police crime figures with victimisation survey data would provide a fuller picture of harm, and help government and the public interpret trends more meaningfully.

Over time, other key datasets could be incorporated, such as mortuary-based homicide records and figures from emergency medical services. Bodies such as the Council for Scientific and Industrial Research and the Human Sciences Research Council could contribute technical expertise to strengthen system design, data capture, and validation.

This approach would reinforce the integrity of the national statistics system, in which the SAPS, Stats SA and other bodies each have clearly defined roles.

The SAPS and Stats SA already have an agreement to enhance the quality and integrity of crime statistics

The SAPS would still collect and analyse crime data for operational purposes. In fact, its analytical capacity should be strengthened. This should include building the Crime Registrar’s capacity to inspect and train police at station level, ensuring better quality collection and capturing of crime data.

Good data and analysis are essential for targeted and timely responses. The national release would provide a neutral overview of broad trends. At the same time, local-level data and analysis would support real-time policing and community safety action.

Releasing crime figures through Stats SA would not absolve the SAPS of its core responsibilities. In fact, it should sharpen accountability for things within its control, including disciplinary processes, expenditure, investigations and public trust. It would also make clear that addressing violence is not just the police’s responsibility.

Implementing a more neutral, robust system will have resource implications. It will require coordination and could mean crime statistics take slightly longer to finalise. This would be a reasonable trade-off if it improves the quality and utility of the data.

The exact institutional arrangement matters less than the principle. Crime data must be released in a way that shows government’s commitment to transparency, fosters public confidence in the police, and promotes a collaborative response to crime.

Exclusive rights to re-publish ISS Today articles have been given to Daily Maverick in South Africa and Premium Times in Nigeria. For media based outside South Africa and Nigeria that want to re-publish articles, or for queries about our re-publishing policy, email us.