of civil society, an autonomous judiciary, and democracy itself, if ‘certain schemes to improve the human condition’ (as Scott described them) based on registration and categorisation are not to fail. The nature of the state and the quality of its institutions determine these outcomes. Registration and identification can obviously function either as an instrument of oppression, up to and including genocide, as it did for the Nazi government of Germany or the Rwandan government in 1994 (Caplan, 2013; Longman, 2001; Piton, 2019); or as an instrument to promote and monitor the realisation of equal rights, as in the post-apartheid government’s repurposing of the registration of ‘natives’ in order to expand access to welfare benefits in South Africa (Breckenridge, 2014). Statistics collected for one purpose can also be used for another, whether with benign or malign intent (Engerman, 2012). It all depends on the institutional context and political orientation of the government and society concerned. Individual registration, rather than social aggregates, can also mark the distinction between a colonial government interested in categories only, and a government more interested in promoting equality of opportunity (Cooper, 2012; Gopinath, 2012). Legal identity, national security and border control It is not surprising that many governments have moved swiftly beyond the development or human rights aspects of ‘legal identity’, to highlight instead the primary importance of national security and border control in implementing the new identification systems. The new technologies of biometric identification are particularly seductive from this perspective. In Kenya, for example, from 2018 the Ministry of the Interior began to implement a new biometric National Integrated Identity Management System (NIIMS), known as the Huduma Namba (service number), to supplement the existing national identity card. The need for the new registration system, intended to provide a ‘single source of truth’ for all personal data (a term derived from information systems design), was justified partly on the grounds of facilitating service delivery, but in large part by national security (Government of Kenya, Ministry of Interior and Coordination of National Government, n.d., 2019; Mugo, 2019). An effort in Nigeria to harmonize multiple identity registers into a single biometric database under the National Identity Management Commission (NIMC), also invoked the need to ’increase security and help fight organized crimes and other illegal activities such as human trafficking, migrant smuggling and fraudulent documents’ (Ibrahim, 2017); the ‘critical need’ for the establishment of an integrated DNA database was also asserted (Government of Nigeria, 2020). Similarly, the contract issued to the biometric company Morpho for the new nationality identity card rolled out in Mauritania from 2011, aimed to combine ‘civil, police and border-control applications… making it possible to search the civil databases to check the identity of people registered in the criminal AFIS [automatic fingerprint identification] system’ (ter Hennepe, 2012). Virtually every country in Africa rolling out a new identification system has a similar statement. This emphasis on security and border control is of course not only from the perspective of ‘developing countries’. The European Commission specifically proposes ‘strengthening third countries' capacity to ensure functioning civil registries and fingerprint or biometrics digitalisation’; not (only) for the purposes of service delivery, but as a means to ‘ensure that third countries fulfil their obligations to readmit their nationals’ (European Commission, 2016). Wealthy states ensured that language on legal identity in the original draft of the 9

Select target paragraph3