52. The sharp decline visible after 2004 is difficult to clarify unambiguously. The most plausible explanation lies in the changed definition of a stateless person in the revised Dutch Nationality Act of 2003.68 Until 1 April 2003 persons whose “nationality cannot be determined” were considered to be stateless under the Dutch Nationality Act. Many of them, as is the case at present, were nonetheless grouped into the category “nationality unknown”, for instance because they could not meet the burden of proof required by the GBA to be registered as a stateless person. Nevertheless, with this broader definition of statelessness, naturally more individuals were registered as stateless than has been the case since April 2003. The sudden drop in 2010 is not due to changes in the policy environment, but a result of modifications in CBS’s data collection methods. In order to acquire its data on statelessness, the CBS now aggregates information provided by municipalities to give an indication of the national state of affairs. According to a CBS demographer, before 2010 the bureau did not receive any information on an individual’s previously registered nationalities. Instead, municipalities provided data on the situation as it prevailed on 1 January of every year. However, as of 1 January 2010, CBS has had access to these “historical” nationalities. Whereas an empty record was previously understood to point to a stateless person, at present this empty record may be complemented by a historical nationality still considered valid today. As a result, significantly fewer persons turn out to be stateless.69 Consequently, data gathered using the new methodology cannot be accurately compared to earlier statistics on statelessness, but is likely to be more accurate than before. The origins of stateless persons in the Netherlands 53. Out of the 2,060 people registered as stateless in 2010, figure 2 shows that a little over 71 per cent (1,471 people to be precise) were born in the Netherlands. Another 13 per cent (260 people) were born in Indonesia, whereas an additional 48 different countries combined make up the remaining 16 per cent (figure 2).70 If we break down the group of Dutch-born stateless persons by their father’s country of birth, the picture portrayed in figure 3 results.71 68  hanges in Dutch migration policy or more proactive deportation efforts are unlikely to be at the C root of this decline, since stateless persons – especially those already registered as such – can in general not be deported and experience great difficulty in leaving the country of their own accord. 69 E-mail correspondence with a CBS demographer, 1 March 2011. On file with UNHCR. 70 CBS, data not available on Statline but on file with UNHCR. 71 Ibid. The eight paternal nationalities depicted in figure 3 represent 76 per cent of the total number of stateless persons born in the Netherlands. The remaining 24 per cent are divided over dozens of other paternal nationalities. Please note, however, that the CBS data do not list the number of persons whose fathers are unknown. At present, the data portray a fictional situation in which every registered stateless person has an identified father. It has been opted to display paternal over maternal nationality in this chapter, because globally nationality laws exhibit a paternal bias. In short, if individuals have a nationality, it is in general most likely to have been passed on by their father. in the Netherlands 23