We should be clear about what we want and what we don't want from a digital ID - about what data is being collected, from whom and how both government and third-party vendors will use it.
In the State of the Nation address last week, President Cyril Ramaphosa promised South Africans a new digital identity system to help "transform the relationship between citizens and government".
Whether the new ID system delivers on this promise depends on how it is designed, implemented and governed. Although digital IDs present genuine opportunities, the risks they pose are just as great, especially in countries such as South Africa, with its extreme inequality, mirrored in high digital inequality.
There is a sense of shared community among those who have stood in a queue at Home Affairs. The institution has become something of a cultural touchstone - a place we go for the collective ritual of bureaucratic frustration.
While a trip to Home Affairs is taxing for most, it is unquestionably worse for those who have never had the required documents, such as a birth certificate, and who are facing economic distress. So, in this light, it is rather exciting to hear that we will receive a new digital ID system that promises to make government services "accessible to every person at a touch".
Foundational digital ID systems convert paper-based legal identities (like South Africans' green identity books, or green mambas) into digital data for processing. They have demonstrated compelling efficiency gains across the globe, including in developing nations. The World Bank, which runs a digital ID for development (ID4D) initiative, estimates that Uganda saves around $7-million a year by using a national identification database to verify the identities of its civil servants. Malawi reportedly saved $44-million by merging its voter registration and national ID systems.
But there are also troubling implications of digital IDs that surpass efficacy gains measured by the dollar, rand and cent.
Some of these issues are inherent to ID systems - the Nazis' census, South Africa's dompas, and Rwandan identity cards serve as stark reminders of the enabling role identification tools can play in systemic oppression. Whether paper-based or digital, ID systems determine where a government or even another party (like a bank or a medical aid company) turns a blind eye and how it uses the knowledge it has to make decisions about a population.
Simply digitalising the system is not a silver bullet solution and indeed may amplify existing inequities while introducing new vulnerabilities.
Of particular concern are the threats of increased surveillance and a loss of accountability raised by some contemporary digital ID implementations. As more of our daily activities become digitised and mediated by technology, each of us is generating more data.
Digital IDs, which can consolidate this data, offer the government increased visibility of its citizens and other residents. This includes greater oversight of the status and activity of historically marginalised populations like migrants and refugees.
Essential services access
Perhaps most troubling is an emerging tendency for digital IDs to become gatekeepers to essential services, while introducing privacy concerns. Across sub-Saharan Africa, an estimated 470 million people lack official identification.
Where digital IDs become prerequisites for accessing healthcare, banking and telecommunications (as was the case in Ghana), such systems risk deepening existing marginalisation.
Uganda's Ndaga Muntu digital ID system offers a cautionary tale here, having reportedly locked out swathes of women, elderly persons and other vulnerable groups from accessing public services. While India's Unique Identity Scheme (Aadhar) has enabled the country to overcome significant identification problems, it has also excluded parts of the population and has raised important concerns over both private and public sector surveillance.
Closer to home, the South African Social Security Agency's introduction of a biometric verification system in its grant system last year mistakenly prevented hundreds of the country's most economically distressed from accessing lifeline funds, purely on account of technical barriers to using the online verification process.
Like all technologies, digital IDs aren't neutral and can come with both promise and peril. The potential outcomes of these technologies are shaped by factors such as how they are designed and implemented and, further, by unresolved underlying colonial histories and foreign aid agendas.
In these complex environments, a growing collection of state and non-state actors also act together to put on a compelling show of technology theatre: a tendency to brandish flashy technology to divert attention from serious policy problems. As yet, we have limited information on South Africa's proposed new system, and we need to make sure we are not setting the scene for another such theatrical display.
As a nation, we should be clear about what we want and what we don't want from a digital ID; about what data is being collected, from whom and how both government and third-party vendors will use it.
South Africa has strong constitutional and administrative justice governance frameworks and a highly regarded Information Regulator. We have the makings of adequate guardrails that can help harness the opportunities and mitigate the risks of a new digital ID system.
The challenge now is to leverage these protections to ensure that digital identification enhances rather than undermines our democratic values.