Last week, I was very lucky to attend a colloquium at the CRC4D in the Hague on ‘The Future of Legal Identity: Implications of New Technology for Civil Registration and Identification’. As someone researching the economies and value chains forming around data in Sub-Saharan Africa, I was particularly interested in understanding the business interests involved in the new biometric systems currently being rolled out across African countries but I ended up coming away with a much broader understanding of the role of identification in development. I thought I would share some of the highlights from the meeting:
1. The Problems of Illegibility
To me, the key weakness of James C Scott’s Seeing Like a State was his proposition that states necessarily want to see their subjects. Yes, surveillance and legibility are forms of state-craft, but so too can illegibility and indifference become part of a power accumulation strategy. To put it simply, if one does not count one’s subjects, one does not have to be accountable for their livelihood or security. If one does not document inequality or corruption, then such problems are less visible to ones people. States may therefore have a disincentive (or just a plain lack of incentive) to document the lives of their citizens. Most African states only register about 40-50% of their citizens at birth. From the point of view of the ordinary woman in the street, a lack of legal identity may make it hard for her to access social services like health or education. She may struggle to vote, get a formal job and or register her business. If she cannot register her business, then she may be subject to arbitrary harassment and informal taxation. She may even find herself stateless if circumstances change and she is forced to move across an international (or otherwise invisible domestic) boundary. From the point of view of power-holders, if 50-60% of your population lacks a legal identity, it becomes difficult to maintain security and to establish more representative forms of governments (in terms of both taxation and social democracy). Just as there are dangers associated with high modernist legibility, so too are there dangers associated with illegibility. In some ways, this problem was the main theme of my PhD thesis (in relation to educational qualifications) and so it was great to be in a workshop where illegibility was the main issue!
2. Parallel Systems
Because so many African states lack the resources and capacity to implement birth registration and vital statistic systems and donors are under pressure to more closely monitor their programs and subsidies, there has been a proliferation of individual functional identity systems, and much less support for establishing strong foundational systems. Donors and development organisations have launched separate initiatives aimed at protecting (and tracking) refugees, delivering (and tracking) social protection and social security and extending financial services to the unbanked. There is a danger that all of these non-state systems are leading to a kind of ‘hollowing out of the state’ and there are of course problems with inefficiency if efforts overlap.
3. Using Digital Traces as Proof of Identity
There were quite a few participants interested in statelessness: where no state recognizes an individual as a citizen and thus the person is left with no legal nationality. This is not my area of expertise but I found the issue of much interest. There is a big international effort to establish forms of ‘legal identity’ separate from citizenship. In fact, guaranteeing legal identity has become one of the SDG targets (16.9). Identity credential would allow vulnerable people to navigate bureaucratic systems in a more legal and transparent way. To some in the group, the proliferation of parallel systems might actually be a blessing in disguise if such systems allow individuals to begin to establish themselves as law-abiding persons with digital traces to prove it. Whether such functional systems could provide the basis for legal identity is an interesting question.
4. High Tech Solutions to Complicated Political Problems?
A very simple conclusion was that while there is a lot of money going into developing high tech ID programs, the real problems lie in solving complicated political and legal problems surrounding identity issues and state-building challenges. Unless you resolve these kinds of political problems, then it is unlikely technology can really help.
5. Understanding Incentives
The question of incentives came up as crucial. Power-holders need to be incentivized to introduce ID systems. Citizens need to be incentivized to register themselves. Private sector actors (apparently) need to be incentivized to provide the services. Therefore, understanding what makes a successful ID system means understanding how these incentives interact with one another.
Incentives for States Elections have proved a powerful motivator for states to introduce biometric systems. Having a free and fair election strengthens state sovereignty and legitimacy of rule- If one wants to look like a modern state these days, one must really have a biometric voter registration! Donors too, apparently want to fund such systems as part of democratization efforts. However, election systems are usually one-offs. A company comes in, provides the service for the election and then leaves again (perhaps taking the personal data with them). In the workshop, there was some talk of how this election-imperative might be harnessed to deliver more long-term identity systems.
Incentives for Citizens South Africa has drastically expanded the registration of its citizens because its grant system is linked to registration at birth. In other countries, it is only when children reach school age that they feel the brunt of not having a legal identity. Anna Rader also talked about how in some countries like Somaliland, there is a deep antipathy to enumerative schemes like censuses. As one of the organizers, Keith Breckenridge said in the workshop, “We’re not going to capture people (into the system) unless they receive an immediate benefit”. There was some discussion about whether the South African ‘grant model’ might provide a template for other countries. However, others warned that raising the stakes of registration also raises the costs of not being registered, and this may be particularly harmful for more vulnerable communities like migrants or orphans.
Incentives for Implementing Parties Because of the high costs of providing such services (and due to the current preference in international development for public-private partnerships), many biometric systems are being delivered by commercial organisations. For example, Sanjay Dharwadker talked about how the Indian biometric system, Aadhaar was broken down into 18-19 different commercial packages in order to make it scalable across the country with numerous private companies delivering the system in different areas under different business models. There was the usual talk of finding ‘innovative financial mechanisms,’ ‘building commercial functionalities into registration systems’ and ‘leveraging’ the private sector to make ID systems ‘sustainable’ in the long run’. However, I think there are potential dangers in this narrow conceptualization of sustainability.
First of all, there may be a trade-off between getting an outside party to implement the system and thus keeping short-term costs down vs. building local capacities and thus potentially reducing costs in the long-term. While some may argue that African states lack the capacity or autonomy to implement and oversee biometric registration systems, state capacity can be built over time. I wondered if the example of autonomous and semi-autonomous Revenue Authorities emerging across African countries might provide an interesting institutional model. Additionally, I was really interested in the case of Aadhaar in India. While private sector companies are implementing the system (not the Indian state), most of these companies are Indian companies and therefore the local economy is presumably benefitting both from jobs and from technological learning and innovation around these systems. This further keeps the costs of implementing these systems down. As so many African governments are trying to build techno-parks and foster ICT innovation within their economies, there may be scope for schemes that mix biometric registration with targeted industrial policy-making.
Second of all, when it comes to the commercial value of ID systems, we need to pay closer attention to value chains around personal data. As was highlighted in the workshop, there are current debates happening in the UK over whether NHS data should be sold to commercial parties. Transport For London is already selling its data to private parties. In Africa, many biometric systems are being integrated into banking systems, potentially providing data about credit-worthiness. It may be very difficult for states to quantify the value of their citizens’ data and to appreciate the varied (and perhaps unexpected) ways that citizen data may be used in the future. In short, as I have been arguing in relation to big data systems, we need a far better understanding of the value chains around data! And as one workshop participant warned, these big ID companies have the power to ‘bully’ the state, further making it hard for African governments to manage the incentive calculation.
6. The South American Experience (States Standing Up?)
Present at the workshop was a wonderful woman called Mia Harbitz who has been advising South American governments about ID systems from within the Inter-American Development Bank. Many South American governments have raised their registration figures from 60% (ish) to almost full coverage. She talked about the importance of advising governments over the technical specifications to make sure that they do not become dependent on one vendor. If anyone is interested in Mia’s work, the IADB website has some useful material.
These are just a few highlights I picked up from the workshop. I noticed Mariana Dahan from the World Bank has also written an interesting blogpost about the event.