Digital identity and biometrics in Africa, Part 1: hopes & promises
This is part of a series summarizing research I did for Dovetail Labs. I had two purposes in writing up this research for a general audience. One, to present an overview that is wide-reaching and complex. In this regard, I went for breadth rather than depth. Two, to identify significant overarching themes in how this technology is being applied, which emerged from an in-depth literature review. Questions of surveillance, state and corporate power, civil liberties, and global inequality were paramount. Three, my goal was to apply the critical approach of academic research for a non-academic audience interested in practical applications of new technologies.
Digital identity in Africa: an overview of political and ethical hazards
This research begins with the recognition that all technology (no matter how apparently “neutral”) has the potential to shape our social and political worlds. New technology affords certain things while it discloses or obviates others. It points us towards certain scenarios and away from others. It embodies assumptions — well-informed, hypothetical, or completely unexamined — about users and their world and in the process, it structures that world. In this regard, technologies have power even before they are picked up as tools by social actors; they are political, whether or not they intend to be.
ID4Africa was a starting point for looking at digital identity in Africa. It was also a case study for fleshing out political and ethical concerns of emerging technologies in non-US settings and a call for ethnographic research on this topic by designers and developers. Digital identity today is a convergence of a number of new technologies in various fields, including digital biometrics, FinTech, and machine learning. It is fascinating in part because, already at its inception, it is linked to disparate sectors, concerns, and phenomena. This case study is an opportunity to think about how to approach omni-use technologies in social, cultural, economic, and political contexts substantively different from those in which these technologies are developed. Whenever possible, I draw on research by people who are either native to the regions they write about, or have spent considerable time studying and thinking critically about them, and I’m partial to researchers who have done field work and/or incorporate on-the-ground observations.
What is ID4Africa?
ID4Africa is a self-described “movement” whose mission is to build “robust national systems that provide legal identity to all Africans.” The organization is technically advocating for legal identity in the narrow sense: documentation that verifies identity and citizenship that is essential to accessing government services and benefits, participating in the democratic process (i.e., voting), and opening a bank account.
ID4Africa was founded in 2014. It is essentially the African arm of the United Nation’s ID2020 project that aims to fulfill the Sustainable Development Goal 16 to provide citizenship and identity documentation to all people.¹ This is an acute problem in sub-Saharan Africa where less than half (46%) of children under 5 have their births registered. ID2020 and ID4Africa cite the UN Convention on the Rights of the Child that states the right to identity as a universal human right.²
This takes place against a background in which, over the last decade, biometric identification systems have become the techno-political model of ultra-modernity in developing countries. Several national biometric identity systems have been rolled out over the last decade. The largest is the Aadhaar identity card in India: as of 2018, 92% of India’s resident population of 1.339 billion have received the card and been registered in this system. There is an expanding range of biometric identification technologies and many possible configurations, each with slightly different implications.
The ID4Africa organization is founded on an explicit tripartite collaboration between three stakeholder groups — African governments, development agencies, and industry (producers and managers of identity infrastructure). The executive committee also has strong professional ties to the global identification and biometrics industries. This partially explains why ID4Africa promotes digital identity as a means to modernize African civil registry, governance, and banking systems, and more specifically, biometric identification.³ In addition, the Board of Advisors comprises members from the World Bank, UNICEF, the Center for Global Development, and the African Development Bank.
My initial research questions
There are a few questions about ID4Africa that I sought to answer, some basic, some complicated. The first was the nature of digital identity. What is it, exactly? How is it intended to be used by developers? The second is the question, basically, of how it is actually being used on the ground and what users think of the technology. There’s very little research done on this; I extrapolated from what I found. Third, I go over the arguments for and against this technology, reviewing the hopes that are invested in it and the concerns that have been expressed about it. Moving towards a conclusion, I ask what might be the significance of this technology in Africa in the context of global markets for financial, identity, and security technologies? What are the potential ramifications of certain private companies cornering the market on digital identity or biometric identification systems in Africa? Lastly, I point to a couple overarching themes and to questions going forward — questions that I think will be relevant for designers, advocates for democracy, and community organizations.
The diversity of Africa notwithstanding, there are a few common conditions that are relevant to the implementation of digital identity on the continent. Many African countries (and other under-developed post-colonial nations) are characterized by weak state bureaucracies, porous borders, and significant informal economies. This can be traced back to the extractive economies and racist policies of colonialism (and, arguably, neoliberal policies that maintain global economic inequality). It is these conditions that are being targeted for improvement by digital identity schemas. Advocates argue that digital identity will radically and beneficially transform these realities. However, there are several critiques of this argument, as I will describe (in Part 2).
What is digital identity? What is biometrics?
Digital identity is narrowly defined as the data profile of a “natural person” (a biologically existing person) that circulates within a given information system, be it commercial or governmental.⁴ The integrity of digital identity as a technological concept rests on authentication and verification: authentication at the time of registration (i.e., making sure the digital data is linked to the correct natural person) and then verification at the time of transaction (i.e., reaffirming this initial data-person link).⁵ Digital identity is getting more attention as a growing number of state services are accessed online, thus making it an increasingly crucial interface between citizens and states.
Biometrics is simply the use of biological measurements (assumed to be permanent) as means of identification. Digital identity does not necessarily involve biometrics. For example, the use of PINs as verification for credit cards and passwords for bank accounts are two forms of digital identity that do not use biometrics. Conversely, biometrics are not necessarily digital. The use of fingerprints by law enforcement and the use of photographs in identity documents (e.g., passports) are also forms of biometric identification. That being said, biometrics and digital identity are becoming both conceptually and functionally intertwined for the simple reason that biometrics is seen as more secure than passwords and PINs, which can be forgotten, stolen, or hacked.⁶ (More on that later.) Biometrics, in contrast, tether the “natural person” to their digital profile through biologically unique information. In the identification industry, biometrics is treated as the cutting edge, and inevitable future, of digital identity.
So that’s digital identity in a narrow sense, but what about digital identity in a wider sense? As I mentioned above, the topic runs in multiple directions because the technology itself is omni-use. As a verification of citizenship, digital identity impacts voting and border control but also much more pedestrian but essential services like health care, education, and subsidies (e.g., food). Some digital identity cards are also intended for use in economic transactions and can be used as a debit card to withdraw cash or pay for purchases. In this function, digital identity touches on how people relate to money, banks, and their national governments as the creator and backer of currency. It acts as an individual’s avatar within political and economic systems that are now mediated by digital technology; it becomes the key to unlocking resources and participation in political and economic processes.
An ethnographic case study in biometric identity: Ghana
Digital identity cards and biometric identification systems have been rolled out already in several African nations and continue to spread across the continent. To introduce this topic, I turn to digital identity in Ghana as a case study.
First, a brief demographic survey. Ghana is a West African country with a population of 25 million. As typical of Africa, it is extremely diverse, with more than seventy indigenous ethnic groups and eighty languages spoken (and English as the lingua franca). Slightly more than half the population is considered urban (as of 2018), two-thirds are literate (as of 2015), and about half of rural Ghanaians live without electricity (as of 2013). A little over a third of the population is estimated to use the Internet (as of 2016).⁷ While Africans in general use mobile phones more than landlines, Ghanaians especially are exuberant consumers of mobile phones.⁸ In fact, according to the CIA Fact Book, there are 134 cell phone subscriptions per 100 people in Ghana. According to a 2016 Pew Research report, 21% of Ghanaians use a smartphone and this number appears to be climbing. All in all, these numbers paint a picture of an ethnically diverse, urbanizing society where high levels of mobile phone use and digital literacy coexist with rural under-development. And since the late 2000s, digital identity has been implemented in Ghana to address election integrity, national registration, and secure banking and payments.
For example, in 2008, near-violent protests took place after a presidential elections was contested by both political parties. Accusations of vote tampering arose due to a high number of spoilt ballots and bloated voter rolls. One solution proposed in the wake of this election was a new voter registry using biometric identification. In an article in the Georgetown Journal of International Affairs in 2010, Gabby Asare Otchere-Darko, the founder and the executive director of a governance think tank in Ghana, argued that a biometric voter registry “would undoubtedly help overcome voter fraud… and would constitute a major step in restoring public faith in the electoral system.”⁹ That being said, he notes the possibility that unfamiliarity with digital systems could simply sow another form of distrust — or as the Chairman of Ghana’s Electoral Commissions put it (in response to a 2010 conference analyzing e-voting models), “We can only go e-voting when my grandmother learns to use the computer.”¹⁰ Oschere-Darko recommends a mixed model using digital fingerprint verification followed by paper ballots so that voters can see the tangible evidence of their vote.
Another example is the Ghana Card, the national identity card of Ghana that allows access to state resources, like schools, and the ability to apply for Ghanaian passport. Between 2009 and 2011, the government began rolling out the new version of the card with biometric and “smart” capabilities. The goal was to capture all citizen data in one centralized database from which governmental agencies could draw, thus streamlining services. In this new iteration, a microchip holds fingerprint, photo, and signature data, accessible through RFID.
Alana Thiel, a political scientist, did post-doctoral research on the implementation of the new Ghana Card in the late 2010s. In her short article on the topic, she documented how the roll-out stalled due to lack of funding and bureaucratic failure. Out of the 15 million Ghanaians registered for the card (in seven out of the country’s ten regions, between 2008 and 2013) only 2.7 million cards were produced and only 1 million distributed.¹¹ In 2013, a new company was contracted to provide the registration technology but that required a new loan from the US government (via the World Bank) as well as the re-registration of citizens, a major undertaking. Thiel documented the apathy and frustration among Ghanaians who registered for the card but never received it. They were looking forward to one card to replace the multiple cards required at the time — one for each government branch with its own separate biometric system, offices, and long lines. Still, the public perception of the Ghana Card remained hopeful. She noted that ethical concerns over biometric identification were only slowly emerging in Ghanaian public discourse.¹²
In addition to Thiel’s work, historian Kieth Breckenridge (in an article on the E-Zwich card; more on that below) also describes two main challenges to Ghana Card registration, both related to the basic question of how to verify who is Ghanaian. As it turned out, only a tiny minority of Ghanaians had birth certificates that could verify their citizenship. The first challenge, then, was that the agencies responsible for the birth and death documents that are needed to verify citizenship had been direly underfunded and understaffed; the second was that, given the fluid boundaries of the state and practices of migration in border regions, the boundaries of citizenship were often blurry. As he wrote:
[T]he Ghanaian state ended up relying on the temporary workers to make a decision about nationality based on discussions with ‘your parents, head of family and/or local traditional leaders’ (NIA, Ghana 2007a). In practice long queues and production pressures meant that those with anything resembling a case for Ghanaian citizenship generally got it.¹³
For the Ghanaian public, all of these realities were further evidence of the need for the Ghana Card. Quoting officials, the card was seen as a means to improve “the integrity of both ‘public and private business transactions,’ and… ‘increase social services.’”¹⁴ Digital identity is seen as a vast improvement on the current state of bureaucracy in the country.
In that same article,“The World’s First Biometric Money: Ghana’s E-Zwich and the Contemporary Influence of South African Biometrics,” published in the journal Africa, Breckenridge looks at the e-Zwich, the biometric money proposed as part of a new interbank switch in Ghana.¹⁵ As a form of biometric money, the e-Zwich is a smartcard that uses a unique key derived from the ten fingerprints of cardholders and encoded into the card. Against the background of post-colonial African attitudes to money, Breckenridge notes the “vernacular logic” of money in West Africa. Historically speaking, attempts by state governments to centralize the control of money in postcolonial Africa have failed, casting doubt on monetary value purportedly created by the state. In short, the vernacular logic of money in West Africa dichotomizes market value, derived from an informal economy, and monetary value as backed by the state authority. If the e-Zwich project does succeed, it will signal a radical departure in the way Africans relate to money.
Alongside these conceptual considerations, Breckenridge also noted a number of logistical problems that were encountered in the card’s implementation. These included scarcity of the point-of-sale (POS) machines needed to read them, problems with finding electricity and Internet connections in rural settings, and user interface confusion.¹⁶
A handful of patterns and issues relevant to the question of digital identity in Africa can be drawn from the Ghanaian case study. These will become more clear after reviewing arguments for and against digital identity in Africa.
Arguments for digital identity in Africa
The basic premise of ID4Africa is that digital identity systems will address obstacles to economic development and good governance in African countries. These can be organized into four main categories: civil registration, access to and expedition of government services, the formalization of the informal economy, and the stabilization of political instability.
The original raison d’être of ID4Africa was, as we recall, as a solution to the lack of civil registrations in Africa, per the UN’s ID2020 campaign. The problem of statelessness is a global one and is much more complex than I could sum up here. The mission of ID4Africa rests on policy recommendations by scholars who have noted the widespread inconsistency, arbitrariness, and ultimate insufficiency of civil registration, which effects access to government services. To understand why this is, we have to start with the fact that many modern borders in Africa were imposed arbitrarily by colonial powers, irrespective of tribal-ethnic land use. For this reason, social and economic patterns are often at odds with modern national identities and boundaries. The most common examples are the grazing routes of nomadic peoples and the tribal-clan networks that transverse borders.
But another effect of this history is that minority groups in countries dominated by a different ethnicity or religion can often face discrimination. Coupled with a lack of due process and weak provisions for birthright citizenship, applications for citizenship (e.g., the civil registration of child) might be denied arbitrarily by administrators. Another iteration of discrimination along these lines is the history of expulsions on the basis of ethnicity and religion in Africa. Nigeria in the early 1980s and Ethiopia in the late 1990s are two well-known examples. These expulsions leave people who had citizenship in one country stateless in another.¹⁷ (History includes much more harrowing examples.)
Low levels of civil registrations are a seemingly mundane manifestation of this history. In this context, the argument for digital identity piggybacks on campaigns for identity documentation led by human rights activists, development professionals, and scholars. It appears that coalitions like ID4Africa have absorbed this human rights rhetoric and then bundled in digital identity as if it is taken for granted. However, the digital-ness of digital identity is based, in fact, on other arguments. Onto those arguments now.
Arguments for digital identity in Africa can be divided into three basic categories: centralizing, and thus expediting, government services; “capturing” the informal economy; and securing insecure political processes, especially voting and border-crossing. All of these can be seen as attempts to solve “leakages” within government infrastructure — of money, services, resources, votes, people. Digital identity, and especially biometric identity, is seen as inherently more secure in that it leaves a trace and, in the case of biometrics, is difficult or impossible to fake. Moreover, it introduces standard and universally recognized identity-verification procedures that will allow for an easier and more secure flow of money between government offices and bank accounts.
In the case of government services, such leakages are better understood as fraud (on the part of beneficiaries) and corruption (on the part of bureaucrats). It is believed that the implementation of digital identity systems will save money in the long run through cutting down on these losses. But it is also believed that digital identification will help centralize government services. This is because national digital identity systems entail one centralized database and (usually) one identification card that will then be used by multiple bureaucratic arms of the government, as we saw in Ghana.
One of the ways that these service leakages would be thwarted by digital identity is by linking government payments, grants, and subsidies to bank accounts through digital flows. What’s more, if the digital identity card is connected to a bank account (e.g., the Nigerian eID card), it can be used as a debit card and capture transactions that might otherwise take place “under the table” (as we say in the US), i.e., that are not reported to the government and thus not taxed. Besides taxes, there is another reason for government to capture these transactions. It is estimated by the International Labor Organization that two-thirds of the total economy of sub-Saharan Africa is informal, which means it’s not accounted for by any governmental measures. From an international development perspective, this means that the ultimate wealth and productivity of a nation or region is vastly under-recorded. This matters to governments courting foreign investors and multinational companies who pay close attention to these numbers in figuring out where and how to invest.
Lastly, one of the main arguments for biometric digital identity in Africa is an answer to insecure borders and elections. The use of biometrics at borders is already widespread, for obvious reasons. As for election integrity, biometrics will arguably guarantee the principle of “one person, one vote.” While there are a handful of undoubtably authoritarian countries in Africa, there are many more that are considered “hybrid” or “illiberal” democracies by political analysts. These are political systems with multiple political parties and seemingly democratic elections but a lack of public oversight over government officials and processes. As a result, a general distrust of the electoral system is common, leading to accusations of election-rigging and voter fraud. Contested election results (e.g., Ghana in 2008; Kenya in 2007 and 2008) are one example. In this context, biometric voting is seen as a way to reinstate trust in the voting process. In one publication, Gelb and Diofasi, researchers at the Center for Global Development in Washington D.C., note that following the 2012 election that used biometric voter registration, in a poll in Ghana found that over 75% of voters polled said that the biometric registration was an improvement over the old system and 87% thought it was “a useful tool of promoting credible and peaceful elections.”¹⁸
In all of these arguments, we can see that digital identity and biometric identification are means to help fortify the workings of the state government and commercial banks. In this regard, supporters argue that this technology will make peoples’ lives easier, safer, and ultimately more plentiful. It appears that many people in Africa believe this argument and trust in the power of this technology to improve their lives.
[Read Part 2 of this series here.]
[1] The UN’s SDGs were introduced in 2015 and are meant to be completed by 2030.
[2] Sullivan, Clare. “Digital identity — From emergent legal concept to new reality.” Computer Law & Security Review 34 (2018): 723–731
[3] The Executive Chairman, Dr. Joseph Atick, has worked in the identity industry since 1998 and co-founded the International Biometrics & Identification Association (IBIA). Greg Pote, the Chief Information Officer, is also Chairman of APSCA, a regional industry association for companies working in payments and identity in Asia, an association he co-founded in 1997.
[4] Sullivan, Clare. “Digital citizenship and the right to digital identity under international law.” Computer Law & Security Review 32 (2016): 474
[5] Sullivan (2018): 724
[6] Fingerprints in bureaucratic archives are considered the first form of biometrics. See: Breckenridge, Kieth. Biometric State: The Global Politics of Identification and Surveillance in South Africa, 1850 to the Present. Cambridge, UK: Cambridge University Press, 2014
[7] Most of these facts are from the Central Intelligence Agency “Factbook.”
[8] Also see this report on the mobile phone market in Ghana.
[9] Otchere-Darko, Gabby Asare. “Ghana’s Fragile Elections: Consolidating African Democracy through E-Voting.” Georgetown Journal of International Affairs 11, no. 2 (2010): 67–73
[10] Otchere-Darko, p. 70
[11] Thiel, Alena. “Entangled temporalities: Ghana’s national biometric identity registration project.” Anthropology Today 33, no. 1 (2017): 3–5
[12] Ibid.
[13] Breckenridge, Kieth. “The World’s First Biometric Money: Ghana’s E-Zwich and the Contemporary Influence of South African Biometrics.” Africa 80, no. 4 (2010): 646
[14] Ibid., p. 647
[15] An interbank switch is the shared mechanism that allows for the settlement of electronic transactions between banks, such as those using ATM cards. This allows for people using local banks to process transactions through non-local network.
[16] Breckenridge (2010)
[17] Manby, Bronwen. “You can’t lose what you haven’t got: citizenship acquisition and loss in Africa.” In The Return of Banishment: Do the New Denationalisation Policies Weaken Citizenship? Edited by Audrey Macklin and Rainer Bauböck, 17–22. Fiesole, Italy: European University Institute, 2015. Also see: Manby, Bronwen. Citizenship Law in Africa: A Comparative Study. Cape Town: African Minds/Open Society Foundations, 2016.
[18] Gelb, Alan and Anna Diofasi. “Biometric Elections in Poor Countries: Wasteful or a Worthwhile Investment? (Working Paper 435)” Washington D.C.: Center for Global Development, 2016.
