UNHCR, Statistical Reporting, and the Representation of Stateless People 91 Yet, to arrive at a better quality of data, the top-down approach of applying narrow legal definitions should be revisited. As noted earlier, most censuses rely on self-identification, while UNHCR’s definition of who counts as a stateless person is determined by the definitions found in the 1954 Convention. In this context, the inclusion of questions on citizenship in national censuses is indicative of a broader problem. While capturing data on those who may be considered de jure stateless, the use of national censuses also offers an opportunity to clarify what is meant by “indeterminate nationality,” and hence evidence of nationality and state recognition. Questions posed to respondents should, therefore, capture data on their access to rights, concomitant with definitions of citizenship. Although this approach takes us well beyond UNHCR’s reach, it would nonetheless assist the agency to have more standardized definitions and to remove some of the exceptions as found in the small print of its official reports. In terms of UNHCR’s own reporting, as Kerwin et al. contend, the different ways in which statelessness arises require specific methodologies that cannot be applied across the board.94 For example, if stateless people are located in a region that has witnessed succession or defederation, then that may encourage an investigation of rates of naturalization among de jure stateless people and the incorporation of those data in subsequent estimates. One might reasonably start by investigating de jure population estimates at the point when new nationality legislation is introduced. Equally, in countries that have introduced new nationality reforms, as in Madagascar where women may now pass on nationality to their children, it would be appropriate to adjust figures. In this instance, official demographic information would record that, under stable conditions, the numbers of stateless people would decline as children reach the age of majority. When adjusting population estimates, it is important to consider the interplay of other factors. If, as Balaton-Chrimes et al. observe, deprivation of nationality reduces the quality of health,95 then this factor should be considered for its impact on life expectancy – though there are also conflicting studies that suggest limits to this approach.96 The relationship between migration and statelessness should also be interrogated on a country by country basis to assess the status of migrant populations whose nationality status may have lapsed during their time spent abroad and who may be at risk of statelessness. For example, Danish nationals born outside Denmark may lose their Danish nationality on attaining the age of twenty-two, unless they apply between the ages of twenty-one and twenty-two to retain it.97 Other countries similarly place restrictions on foreign-born nationals living abroad. Such an 94 Ibid. S. Balaton-Chrimes, B. K. Blitz, M. Lynch, and R. W. D. Lakshman, The Cost of Statelessness: A Livelihoods Analysis (Washington, DC: U.S. Department of State, 2011). 96 L. Liu and G. Singh, “Mortality Trends and Differentials by Nativity Status in the United States” (2018) 28 European Journal of Public Health 21,https://doi.org/10.1093/eurpub/cky047.008. 97 Ministry of Foreign Affairs of Denmark, “Danish Nationals Born Outside Denmark and the 22-Year Rule,” 2020, https://storbritannien.um.dk/en/travel-and-residence/family-and-legalmatters/dual-and-multiple-nationality/danish-nationals-born-abroad/. 95 https://doi.org/10.1017/9781108914994 Published online by Cambridge University Press