Killing the Files: IPID’s cover-up of police brutality in South Africa
Category: Best data-driven reporting (small and large newsrooms)
Country/area: South Africa
Organisation: Viewfinder, Daily Maverick, GroundUp, Checkpoint on eNCA
Organisation size: Small
Publication date: 10 Jul 2019
Credit: Daneel Knoetze, Timothy Gabb, Alex Noble, Anton Scholtz, Laura Grant, Tomas Knoetze
Decades after apartheid, police brutality remains pervasive in South Africa. The perpetrators are almost never held accountable. Why? This project exposed that SA’s police watchdog, the IPID, had a long history of “completing” cases without proper investigation. This was done to inflate performance statistics, while obstructing justice for victims. In reporting this story, I accessed IPID’s database for more than 36,000 criminal complaints against the police. We analysed the progression of these cases. Whistleblowers, leaks, research and the experiences of victims augmented the data analysis to reveal the reasons, extent and true human cost of IPID’s statistical manipulation.
This project launched with the broadcast of a video documentary on a national television broadcaster (Checkpoint on eTV); the publication of a long-form exposé in partnership with the Daily Maverick and GroundUp; and, a series of newscasts on Eyewitness News’ radio service. On launch day, we approved requests by News24 and IOL – two of SA’s largest online news sites – to syndicate the exposé in full. Through the launch week, our exposé was picked up as a hard news item by various other newsrooms and current affairs shows on national radio and television.
Civil society and policing sector experts immediately called on IPID to account. Notably, the Institute for Security Studies (ISS) condemned the cover-up and called for an independent inquiry. The African Policing Civilian Oversight Forum (APCOF) scheduled a roundtable discussion for sector experts on IPID’s role and challenges. APCOF is in the process of finalising a policy paper based on Viewfinder’s findings. We understand that it contains recommendations for improvements in case management and oversight mechanisms at IPID.
In the midst of this public scrutiny, IPID presented its 2018/19 annual report to Parliament. That presentation was overshadowed by our findings of statistical inflation contained in successive IPID annual reports. MPs called on IPID’s management to account and committed to more closely monitor IPID’s progress on addressing statistical manipulation. Weeks later IPID presented to Parliament again. It acknowledged that some degree of statistical manipulation did occur.
Meanwhile, Viewfinder established a reporting line for victims of police brutality who felt that their cases were poorly handled. We are in the process of assisting numerous such victims to access their dockets from IPID and to analyse whether those investigations were properly done or covered-up.
The first technique I used was to submit a Promotion of Access to Information Act (PAIA) request to the Independent Police Investigative Directorate (IPID) for access to their Master Register databases. My request was granted. I also used OCR to lift d Then, I used two investigative techniques to develop a clear understanding of what the data represented and how the data was produced and compiled: i) I studied IPID’s Standard Operating Procedures, specifically as it related to the progression of cases on the computer case management system ii) I interviewed informants and whistleblowers to understand the institutional environment and pressures which informed the digital aspects of case management and data compilation. Through this latter process I learnt more about how cases were manipulated to inflate performance statistics. Then I worked with data scientists to clean and align dozens of spreadsheets so that they may be compiled into one mastermaster database which allowed us to track the progression of each of more than 36,000 criminal complaints against the police as these progressed through IPID’s case management system. This was a wholly collaborative process, which I directed on the basis of my knowledge about IPID’s processes and what the data at each stage of the case management process. We variously used Python and R to do this cleaning and compilation. Then, with my direction, we analysed the data to identify various red-flags and test various hypotheses: that the conviction / accountability ratio for these crimes was very low, that case “completion” indeed happened at or near performance reporting deadlines (whistleblowers claimed many such cases were poorly investigated), that certain cases were closed or completed in contravention to SOPs. Then I employed the services of a motion graphics professional to visualize certain elements of the data for inclusion in my film and
What was the hardest part of this project?
Viewfinder was founded on the understanding that “public interest” journalism in post-apartheid South Africa should be rooted in the imperative for greater equality and redress in our society. We are creating a space for long term investigations into abuses of power which impact directly on the poor and marginalized majority of South Africans. We are producing journalism which engages audiences across the class spectrum and empowers them with the information and knowledge to hold large institutions accountable on such abuses. Such a dual mandate has been sorely missing from other SA newsrooms in general, and investigative journalism outfits in particular.
This project was based in an investigation that took longer than a year to report. On a shoestring budget and with minimal other support, we collated and analyzed scores of spreadsheets; submitted more than a dozen access to information requests; trawled through large troves of public records; obtained leaks; produced a 25 minute documentary film and traveled to remote parts of the country to speak with whistleblowers and victims of police brutality. We committed a substantial part of our funding to fact-checking and legal vetting – to ensure watertight accuracy and credibility. The result, I believe, is a demonstration of what is possible when bringing best practice in data analysis, conventional investigative techniques, research, solutions orientated journalism, field reporting and storytelling to bear on abuses of power which impact on the poor and marginalized majority in SA.
The hardest part of this project was to keep the faith that we could achieve that objective, over fourteen months of arduous investigation and preparation.
What can others learn from this project?
As a former news journalist committed to reporting on social justice struggles, I used to believe that data-journalism somehow existed in a different realm from what I was doing. For me, as I suspect it is for many other reporters, data journalism was an opaque world of infographics and fancy tools produced by people with different skillsets and interests to myself.
In reporting this project, those assumptions were exploded. My project was less a “data-journalism” project (in the way I had previously conceived such to be) and more a project of investigative and explanatory reporting which “stood on a foundation of data and analysis”. This was a revelation, and one that I would like to share with other reporters that have not yet seen the potential of data as a legitimate and important source for bolstering accountability and investigative journalism projects.
As a reporter I have always valued the importance of good storytelling to engage one’s audience. I think that my project may demonstrate to others the potential of data as a source of narrative inspiration. Data is often seen as depersonalized, neutral and dry. My reporting demonstrated the human story behind the data and demystified the processes which act upon the people implicated in the data and the management of thereof. In my story, the reader is reminded that the numbers represent real victims of rape, torture, murder or assault by police officers. They represent real perpetrators, who often act with unchecked impunity. The premature “completion” of cases is the outcome of decisions and processes by IPID investigators (real people) who are driven to cover-up due to the incredible constraints and pressures of their work.