Open Courts provides the public with access to high court records from around South Africa
We all have the right of access to the courts, to observe how matters will be decided, and to obtain court documents relating to those cases.
The South African court system sees thousands of cases every year and each of these is made up of hundreds, if not thousands, of pages of records. Despite the importance of these records and access to the courts in general, they are mostly only made available in paper format. Part of this is because the court process itself is still largely done on paper, from filings of each of the parties involved in a case, to directions issued by the courts themselves. This can make getting a hold of them difficult, and leaves them vulnerable to loss or decay. The court process itself is slowly changing to become more digital.
Under the banner of openness, transparency and access for all, Open Courts aimed to tackle these challenges and improve access to justice in South Africa.
"If you want to understand a process, let your users create the baseline."
Open Courts is a database of digitised high court records. To create this repository, increase public access to records and promote transparency, and to advocate for digitisation of records across the country, the team (consisting of OpenUp staff members, and partners from Ndifuna Ukwazi and Reclaim the City) needed to (a) understand the needs of those who use court records, both internally and externally, and (b) understand the process of accessing court records.
We started out by identifying different types of people who use court records, including legal professionals, members of civil society, journalists and researchers. This helped to inform the design and technology of the end product. It also raised questions around making current versus closed matters available; how to accommodate differing levels of legal literacy; the accuracy of the scanning process (in this case, optical character recognition, or OCR, is used); how much of a single record needs to be made available; and the differences in the types of cases the majority of users are interested in.
Court records are legally required to be open to the public, with some exceptions. In practice, to access most court records, you will physically need to walk into the relevant courthouse and track down the right person to ask. This can easily lead to getting lost, feeling overwhelmed, and digging through endless records to find the case you're looking for.
While rules of court provide guidance on the access process, each court implements the given process somewhat differently. To try and understand the remits of each courts process, we used three key methods to generate our own process data (on top of desktop research methods):
This experience of testing out the current system helped to further inform the direction Open Courts would take. Generally, the process was prohibitively expensive and challenging: records must be copied inside the courts at the prescribed costs (this is true of some courts but not all of them), and in practise was not well designed for members of the public to try and navigate. This does not speak to openness and proper access, as the law demands.
It is important to remember that many of the courts are under-staffed, overburdened, and lack resources and the additional support required to digitise every single court record. So, in creating Open Courts, the reality of the courts' context needed to be kept in mind.
Once we identified possible users of court records, we needed to try and innovate the existing process. This would adapt and develop a digitisation process, resulting in a repository of records available online. We selected a team of six researchers and assigned them each a case list of nine files to collect, from the Cape Town Magistrate Court. The collection process was staggered so as not to overburden the offices at Ndifuna Ukwazi (where the researchers did their scanning) and after two weeks, a group debriefing took place.
The process turned out to be very successful, especially given the experiences of collecting records from the initial research phase. A total of 52 files were collected and scanned by the six researchers. Researchers were able to collect more than three files per day; re-requests of certain files proved successful on second and third attempts; court staff were helpful and did not hinder the process; and the 'official' process with waiting periods and request limitations were not set in stone.
Overall, the researchers worked well as a team and used a group chat to coordinate their efforts over the two week period. This made them feel continuously encouraged and advice could easily be passed on. Their experiences of collating court records was a mostly positive one and informed the process that Open Courts was trying to create.
In the end, five rounds of record collection took place and with each subsequent wave of accessing and digitising records successfully, the cost per file was reduced. Our efficiency also improved over time and the last round of collection, where 34 records were accessed and digitised, took less than a week. Interestingly, the researchers - who had little to no experience with the courts - found the process far less challenging than expected. They were not given the runaround and almost all of the records they requested were available (even if this took more than one request attempt).
After scanning all these records, we needed to make them openly accessible. To do this, we built a site at www.opencourts.org.za, which not only provided open access, but also helped ensure that they were searchable. Importantly too, it can cross reference characters between datasets.
At its core, Open Courts was created to increase access to court records, especially to the public. It can also be used by researchers and journalists who are looking for trends and patterns in the court records. The need to simply make records open and more easily available was made clear through our experiences of learning how the process currently works, and how it can be improved on going forward. This has resulted in a basic database of several hundred court records that can easily be searched and downloaded in PDF format. If we were to move to the next phase of the project, we would likely develop the tech to include stronger indexing in relation to case types and thematic issues; more records; and functionality that allows external users to upload records.
This blog is based on a report compiled by OpenUp's Gabriella Razzano