Machine Readable Me

Chapter 1: How does data try to capture who we are?

In 2012, I read a tweet from the United Nations High Commission for Refugees (UNHCR), one of the world's biggest refugee agencies. The tweet described the reaction of a Senegalese refugee who had just received an identification card, issued by UNHCR.

'At least I have an identity now. I exist.'4

The accompanying article went on to extol the ways in which gaining an identification card changed lives, allowing recipients to integrate into local society, gaining access to loans and local schools for their children.5 It also noted that the cards weren't just simple identification cards, they also held the owners' fingerprints, photo and biographical data. Why, I wondered, would a card need that much information? And what might happen if it fell into the wrong hands? Those musings, and that statement of 'existing' in large part due to that card, stayed with me and ended up being the spark that shaped my research over the coming years.

The way that data is organised, via categories and labels, can have a huge impact on our lives. Journalist Lena Groeger writes, 'decisions about how to design a form have all kinds of hidden consequences', citing many examples of how form design affected important data collection, such as data about different races gathered via the census in the United States, even influencing who people are more likely to vote for.6 Category creation and curation has always been the source of a great deal of power, long before digital technology spread it faster and further than we'd ever imagined.

In the 1920s, Belgian colonial powers took it upon themselves to institutionalise racial categories that had not, up until that point, played a significant role in Rwandan society. Key to that was carrying out a census that 'classified the entire population as Tutsi, Hutu, or Twa, and issued each person with a card proclaiming his or her official identity',7 before they moved on to reforming local administration according to these new racialising categories, as Ugandan scholar Mahmood Mamdani explores in his book, When Victims Become Killers.

The Belgians, and other colonising powers before them, understood the power of drawing lines in society where there were previously none, and of adding levels of bureaucracy to make life easier for the few that were in power. Under their rule, 'the colonial power constructed the Tutsi as nonindigenous and the Hutu as indigenous . This had a crucial social effect: neither kwihutura (the social rise of an individual Hutu to the status of a Tutsi) nor gucupira (the social fall from a Tutsi to a Hutu status) was any longer possible. For the first time in the history of the state of Rwanda, the identities "Tutsi" and "Hutu" held permanently. They were frozen.'8

Ghanaian-American philosopher and academic Kwame Anthony Appiah calls this the 'Medusa Synd-rome', writing that 'what the state gazes upon, it tends to turn to stone.'9 He describes this inadequate but somewhat inevitable strategy that the nation-state adopts as the only way a state has of making its people legible or, in other words, of 'watching' its population. But watching is not the same as seeing.

Nowadays, our identities are codified in a wide range of systems which are primarily controlled by large institutions through documents like passports and ID cards. States are one of the main players in the identification game, using data to discern those worthy of welfare payments, to charge taxes, to offer health services and pay pensions, and this is nothing new. What is new is the digitisation of these systems. With the support and financial encouragement of the World Bank, more and more governments are setting up digital ID schemes that are increasingly tied to welfare and other government-run programs, further integrated with private sector services.

Before we get too far into the problems of such systems, it's worth understanding why these systems are established on such a huge scale. Back in 2015, member countries of the United Nations agreed upon a set of goals to shape the following fifteen years, which have come to be known as the Sustainable Development Goals (SDGs). These goals have set a framework for a lot of governmental spending since then and include important (though lofty) targets like achieving zero poverty, zero hunger, and gender equality.

Goal 16.9 states that that everyone should have access to a legal identity - that is, a legally recognised document which provides an official record of the existence of that person and enable the recognition of that person before the law.10 Right now, millions don't have that, for a variety of reasons. They might not have access to a mechanism to register births, which means that when they are adults they can't officially prove that they were born. Or they might not be recognised as equal to others in their country simply because of the family they were born into, or the conditions of their birth, and as a result, not be entitled to a legal identity document. This often happens to people born into ethnic minority communities who are discriminated against by those in power - as we'll see shortly, communities like Nubians in Kenya, or Rohingya in Myanmar.

The issue of access to legal documents is important enough that in June 2023, stateless and undocumented people living in the Kavango East and Kavango West regions of Namibia held a demonstration to call on the Namibian government to speed up the registration process.11 The group flagged the irony that death certificates of undocumented people are processed faster than documentation to help stateless and undocumented people live their lives.

Theoretically, digitising this process should mean registration happens efficiently, and be easily recorded and accessed at scale, potentially addressing many of the hurdles towards providing legal identity documents to people. But this is a fallacy. Legal identity is often withheld by governments not for a lack of proper system, but because of societal bias or discrimination. Any digital identification systems that are established without actively addressing those biases will simply replicate that discrimination, thus not really addressing the core issue at all.

We can see this kind of discrimination replication in how members of the minority Nubian community in Kenya have been treated. Their ancestors were forcibly brought from Sudan by the British colonial government over a century ago, and despite there now being over 100,000 Nubians in Kenya,12 they've faced discrimination ever since, with the Kenyan government denying them their rights as citizens of Kenya. They've struggled to access identity documents which are 'vital tools in fighting for other rights such as education, land rights, seeking loans and employment opportunities.'13 Organisations like the Nubian Rights Forum have been doing crucial work helping Nubian people access identity documents in the face of that discrimination, via radio talk shows on citizenship and related issues, and legal support, aiming to empower Nubian community members to know and claim their rights.

In January 2019, the government of Kenya launched a national, mandatory digital identity card system, popularly known as Huduma Namba (meaning 'service number' in Swahili) which would be a prerequisite to access vital social and governmental services. To enroll for Huduma Namba, though, identification documents are needed, documents that many members of the Nubian community struggle to access in the first place. Human rights groups such as the Nubian Rights Forum, the Kenya Human Rights Commission, and the Kenya National Commission on Human Rights have launched multiple legal challenges against the proposed system.14

That's not the only issue with the Huduma Namba system as it was proposed, but it does highlight just one way that digital systems can unintentionally facilitate discrimination. There was enough resistance to the system that in early 2023, the budget for the Huduma Namba was cut by 84%, effectively ending the rollout.15 Uptake was low and rights groups carried out successful legal challenges against the system. Even in the face of well-funded, nationwide systems like the Huduma Namba, well-organised resistance and protest can make a difference.

Digital identification systems such as the Huduma Namba aren't necessarily a part of reaching Goal 16.9, despite its appearance. Legal identity documents are important because they enable the recognition of that person before the law, which facilitates the claiming of other rights. But many digital systems are a prerequisite to accessing government services, as with Huduma Namba - meaning that their actual use is far beyond simply providing legal identity. Often these other government services involve some level of involvement with the private sector, too. When that's the case, they often gather much more data that wouldn't at all be necessary for simple legal identity. For example, the Huduma Namba was originally proposed to include DNA, retina scans, iris patterns, voice waves and even earlobe geometry, all to be stored with personal information in a central database.16...