Digital identity
Digital identity
The World Bank estimates that over 1 billion people worldwide are unable to provide identification proving who they are. Digital identity is essentially a person’s electronic fingerprint — their birth registration, vaccinations, educational certifications, and legal status, all stored online where it can’t get lost.
We have joined ID2020, a global alliance that includes Microsoft, GAHI, Ideo.org, Simprints, and a variety of other actors to help promote principles of “good” digital identity.
In Liberia, national biometric ID cards are already in use. Citizens are required to use their biometric identification (in the form of retina scans, fingerprints, or facial recognition) to open a bank account, register to vote, and obtain a driver's license.
Identification in a digital age matters because it reduces complex humans to records and systems, mostly categorised by others. Consider the refugee who is given a ‘refugee identity’ by UNHCR, or a female platform worker who is identified through a scan of the national ID she submits along with her profile. The term digital identity indicates the conversion of human identities into machine-readable digital data. As noted by Nyst et al. (2016, pp. 8–9), such a conversion results in schemes that allow the digital identification of individuals, as well as their authentication at various points of access and, on that basis, authorisation for them to perform given actions or access given services. Digital identity schemes are schemes in which ‘the three functions of identification, authentication and authorisation are all performed digitally’ (Nyst et al., 2016): such schemes are hence different, for example, from schemes where the delivery of goods or services is digital, but the identification of recipients depends on paper-based identity cards or other documents.
The relation between digital identity and socio-economic development has been made explicit across multiple works (e.g. Caribou Digital, 2017; Dahan & Gelb, 2015; Gelb & Clark, 2013a, 2013b; Gelb & Metz, 2018; World Bank, 2016). While diverse links are suggested, a common basis lies in the recognition of an identity gap (Gelb & Clark, 2013a; UNICEF, 2013), which leaves people – and especially people in vulnerable countries and regions – uncovered by civil registration. While lack of a legal identity results in denial of essential rights (Nyst et al., 2016), its recognition provides the basis for individuals to be entitled with rights, receive public services, or benefit from much-needed forms of social assistance. As a result, the United Nations Sustainable Development Goal target 16.9 is framed in terms of reaching a ‘free and universal legal identity, including birth registration, by 2030.’
Within the broader picture of a link between legal identity and essential rights, digital identity emerges as a means putting technology at the service of socio-economic development (World Bank, 2016). In its diverse manifestations (Gelb & Metz, 2018), digital identity uses technology to guarantee the legal identity that makes people visible to providers as holders of rights, whether they are undocumented poor in urban slums or displaced persons fleeing war and persecution. As noted by the World Bank (2016, p. 2), identification systems have three overarching development goals, framed as (a) inclusion and access to essential services (such as healthcare, education, voting rights, financial services and social safety nets); (b) effective, efficient and transparent administration of public services; and (c) greater accuracy in measuring progress on key development indicators, e.g. reductions in maternal and infant mortality. By strengthening existing identity systems and developing new, reliable ones where these are absent or weak, digital technology affords the possibility to build development through robust identification.
And yet, the so-constituted orthodoxy of ‘digital identity for development’ presents multiple issues.
Research highlights the importance of unpacking links between benefits attributed to digital identity schemes and their costs, firstly in terms of exclusion of legally entitled recipients (cf. Drèze et al., 2017; Khera, 2019; Muralidharan et al., 2016, 2020; Ramanathan, 2014). Determined at the stages of registration, authentication or assertion of identity (Hosein & Whitley, 2019), exclusions result in denial of essential services and may be associated to highly grave outcomes. In India, hunger deaths have been associated to digitally enforced exclusion from food security schemes (cf. Singh, 2019). In Kenya, migrants and refugees are barred from essential social assistance as a result of misrecognised IDs (Weitzberg, 2020). In addition, there is disproportionate impact on women because they often feel (heightened by socio-cultural norms) that they cannot justify the need for ID as compared to men (Bailur et al., 2019). This is particularly problematic in the case of digital gig economy work, which is seen as attractive for women, but where analogue infrastructures may not support their online work. For example, they may be paid for their work online but may not have a bank account or mobile money account, resulting into money being paid into their husband’s account (Caribou Digital, 2020).
There is also a component of how identification (a static process) captures one’s dynamic identity in a fairly immutable way which makes it much more challenging to change later. This is problematic for all, but impacts on those whose identity necessarily is fluid at the time, such as children, refugees or migrants. Caribou Digital’s research with UNICEF on children, identification and identity in Brazil, Kenya, Lebanon and Thailand illustrates how children and youth, born into identification systems, may not understand or struggle with fixed identities, exacerbated in the digital age (Bailur & Smertnik, 2020). They illustrate this with the case of Paul, in Kenya, who uses his brother’s national ID and M-Pesa account to get a small loan from the mobile loan app Tala but who is not aware of the repercussions for his brother if he does not pay that loan (Smertnik & Bailur, 2020a). Similarly, in Brazil, youth struggling to change their name and sexuality on their birth certificate and other credentials face numerous obstacles (Bailur et al., 2020), and young Syrian refugees in Lebanon feel their identity is so much more than a ‘refugee,’ but feel their identity as captured as such in digital identity systems only shows one side of them (Smertnik & Bailur, 2020b). Thus, identification while enabling human freedom in one sense can also restrict it, depending on who is doing the identifying.
For migrants and refugees, being labelled is problematic. Caribou Digital’s research in India (Caribou Digital, 2017) illustrates how migrants within India struggle with their ‘fixed identification’ when moving between states, a challenge that was further apparent during the COVID-19 crisis in 2020 when migrants with ‘Mumbai’ as residence on Aadhaar cards could not take certain trains back to their home states because they were not considered as belonging to those states (Vaktania & Sharma, 2020). Janmyr and Mourad (2018) illustrate how Syrians fleeing to Lebanon resent labels are imposed on them of ‘registered refugee,’ ‘labourer’ or other exogenously determined categories, detailing how such labels impact on them. This runs the risk that increasingly open digital platforms for identification may lead to the monopolisation of identification categories through network effects that force more stakeholders to use the same static categories, when these are actually fluid categories (Janmyr & Mourad, 2018). This is evident particularly in the link of identification to financial sectors, such as credit ratings, where a user may not be in control of what information is revealed about them or be able to explain it. The Black Mirror episode Nosedive always comes to mind as an illustration of how a woman’s life unravels because she is rated by strangers on everything she does.