from office or to justify mass expulsions (Manby, 2016). The African Court on Human and Peoples’ Rights has affirmed the right to a nationality as part of customary international law, and found that Tanzania’s exclusion of court review in questions of immigration and citizenship rendered executive decisions arbitrary (Manby, 2019). Comprehensive guidance has been provided on the interpretation of states’ obligations in the matter of birth registration and the right to a nationality, emphasising the importance of protections against statelessness and the creation of at least some rights to nationality based on birth in the territory, especially over multiple generations (African Committee of Experts on the Rights and Welfare of the Child, 2014). These and other normative endeavours provide the foundation for a more robust insistence – by citizens, by parliamentarians, by civil society, by development agencies, by donors – on human rights principles during the roll-out of new identification systems. This includes especially the right to due process in determining entitlement to recognition of nationality or another legal status; but also moves to reduce various forms of discrimination in nationality administration, and recognition in at least some contexts of the right to nationality in the country of birth and long-term residence. Of course, these are highly political and politicised questions. Yet Africa’s citizens are more open to the idea of rights to citizenship based on birth in the territory and a contribution to society than their politicians have been. Public opinion polling by Afrobarometer in 29 countries in 2011-13 found that almost everywhere a majority supported the grant of citizenship to the child born in the country of two foreign parents, with an average of 59 percent in favour; an average of 62 percent thought a person should have the right to become a citizen based on a contribution through living and working in the country (Afrobarometer, 2013). It remains the case, amidst all these initiatives, that a strategy of ‘state-avoidance’ (Scott, 2009, pp. 30–31 and passim) may be desired and followed by some; notably those of deep rural or nomadic lifestyles, but also those living in urban slums who are subjected only to the coercive capacity of the state and see little of the benefits of inclusion. This strategy is increasingly impossible to adopt with any success (Scott, 2009, pp. 324–325). If it is argued that ‘[u]nder capitalism, the only thing worse than being exploited is not being exploited’ (Denning, 2016, p. 273), we could say something similar about identification systems in a world order based on sovereign states. Even in the most repressive states, it may be the case that the only thing worse than being officially registered is not being registered. But in most states, even those of quite weak capacity, recognition and registration – as a citizen – carries significant benefits. Conclusion: first, do no harm It was ‘the possession and use by citizens (not the central state) of an identity registration system’ that was highlighted by Simon Szreter as the key factor in economic development in early modern Britain (Szreter, 2007, p. 72). Among the policy lessons he drew from this example were that investment in a system of identity registration should be conditional on a democratic constitution and robust civil society, ‘an accessible, impartial, and independent system of recourse to law for the common people’, and management by an autonomous institution (Szreter, 2007, p. 80). Largescale systems of registration are also more likely to work effectively for the state if they are designed to provide an obvious benefit to the individuals registered (Szreter & Breckenridge, 2012, p. 16). 17