The new GLOBALCIT Citizenship Law Dataset systematises information on the different ways in which citizenship can be acquired and lost around the world. The Dataset covers legislation in force on 1 January 2020 in 190 states and is organised around a comprehensive typology of modes of acquisition and loss of citizenship, capturing 28 ways in which citizenship can be acquired and 15 ways in which it can be lost. The new Dataset facilitates comparisons of the rules applicable to similar groups across countries and provides a global outlook on the current state of citizenship law. and the Dataset is designed for use across academic disciplines, as well as by practitioners and policymakers.
In this short introductory blogpost, we explain a) why we needed a new Dataset, b) how the Dataset has been compiled, and c) some of the different ways in which the Dataset can be used (for a longer introduction to the comparative methodology and a more elaborate presentation of some major patterns, see here).
Why a new Dataset?
The previous GLOBALCIT databases on Modes of Acquisition and Loss of Citizenship provided users with qualitative information on relevant provisions in citizenship legislation in 175 countries as of 2016, categorised using a typology of so-called ‘modes’ that facilitated the comparison of rules applicable to the acquisition and loss of citizenship by similar target groups across countries. For example, mode A01a covered legal provisions applicable to the descent-based acquisition of citizenship by children born in the country of the respective citizenship, mode A01b covered similar rules applicable to those born abroad, and so on. This comparative typology was initially developed in the context of 15 European Union countries and subsequently fine-tuned and systematised to allow for global comparison of citizenship laws (see here for a discussion of the application of the functional comparative method).
The new GLOBALCIT Citizenship Law Dataset updates and expands the existing ‘modes’ databases in four respects. First, the geographical coverage of the Dataset has been expanded and now includes 190 independent states, including micro-states from Polynesia that were previously not covered (see comparative regional report by Anna Dziedzic here). Second, all information has been updated and now covers the legal situation as of 1 January 2020. Third, whereas the previous version of the database on modes of acquisition of citizenship included only summarised information for a single A06 mode on residence-based citizenship acquisition, the new Dataset (following a similar strategy to GLOBALCIT’s CITLAW indicators) provides specific data for selected requirements for residence-based citizenship acquisition, such as residence requirements (A06a) or language conditions (A06c).
All in all, these changes mean that the new Dataset has 9310 country-mode entries, covering 34 acquisition modes and 15 loss modes in 190 countries in 2020 (see Codebook for complete overview of modes covered within this revised comparative typology).
Fourth, in addition to qualitative summary descriptions of the relevant legal provisions on the acquisition and loss of citizenship (including references to the relevant legislative provision), the new Dataset now also includes a coding scheme that identifies the main categorical distinctions between legal provisions across countries. Hence, in contrast with the previous version of the database, users can now identify and analyse patterns in terms of groupings of countries or focus on particular distinctions between countries.
How has the Dataset been compiled?
The Dataset covers 190 states and reflects each state’s citizenship legislation as in force on 1 January 2020. The input for the Dataset was principally provided by GLOBALCIT Country Experts or GLOBALICT Regional Coordinators, who were selected on the basis of their specialist knowledge of citizenship policies in their country or region of expertise, as attested by their (academic) research or policy engagement. If a particular country was not covered by either a GLOBALCIT Country Expert or a Regional Coordinator, data were primarily retrieved from available GLOBALCIT sources, including country reports and (translations of) national legislation available in the GLOBALCIT Global Nationality Law Database.
An important caveat is that the dataset is based on the substantive requirements as set out in legislation. In practice, the implementation of the law may deviate from these formal requirements. If it is known that practical implementation deviates from the substantive legal requirements for a given entry, this is indicated in the ‘conditions’ column. For each entry, the dataset records a standardised procedure under which the acquisition or loss of citizenship takes place, but it does not record all procedural requirements, such as the applicable fees, which can be difficult to ascertain and frequently changed.
What can you do with the new Dataset?
We now highlight four ways in which the systematised data on modes of acquisition and loss of citizenship from the new Dataset can be used for descriptive analysis.
First, as was the case previously, the updated data for 2020 can be explored through two online databases on modes of acquisition of citizenship and modes of loss of citizenship. These databases allow two types of search: a) list all modes of acquisition or loss of citizenship in one country (select the country you are interested in); or b) compare a mode of acquisition or loss of citizenship across countries (select the mode of you are interested in and compare across all or selected countries). Users can download the whole datafiles or selected data through the data export function of the databases.
Second, users can also explore the country-year datafile posted in the repository, which includes the quantitative coding outlined in the Codebook. For example, the binary versions of the variables in the Dataset (A01a_bin for mode A01a, A01b_bin for mode A01b, etc) can be used for a ‘big picture’ description of the frequency of the number of applicable modes of acquisition of citizenship across countries. This is visualised in Figure 1, which shows that most countries have ten or eleven modes of acquisition, out of a total of 28 ‘main’ modes (excluding modes A06a-A06f). Nonetheless, as is clear from the normal distribution shown in Figure 1, there is considerable variation. For example, Uruguay has only two birthright-based acquisition modes in its citizenship law (A01b, A02a), whereas in Greece up to 24 out of the 28 acquisition modes are applicable.
Only one acquisition mode (A01b, citizenship acquisition by a person who is born abroad to a parent who is a citizen of that country) is applicable in all 190 countries. Other common acquisition modes include those for residence-based acquisition (A06), marriage-based naturalisation (A08), citizenship acquisition by a person born in the country to a parent who is a citizen of that country (A01a) and reacquisition of citizenship (A16).
When looking at the ‘big picture’ for the modes of loss of citizenship (based on the 15 possible modes), we find that the median number of applicable modes of loss of citizenship is five. The most common are voluntary renunciation of citizenship (L01, applicable in 162 out of 190 countries) and loss of citizenship on the basis of fraudulent acquisition (L09, in 154 countries). The least frequent mode is loss of citizenship by a person whose descent from a citizen has been annulled (L13a, only present in nine countries).
There is also considerable variation, with two countries (Ecuador and Uruguay) having no formal mechanism by which citizenship can be renounced or lost. In the case of Uruguay, which also has the fewest modes of acquisition (two), this suggests that citizenship is regulated only to a very limited extent. By contrast, in three countries (Indonesia, Lithuania, and the Netherlands) almost all applicable modes of loss (eleven out of 15) are present.
Third, besides the number of modes, the new Dataset now also allows users to identify the relevant type of conditions for the acquisition or loss of citizenship, where a given mode is applicable. The frequencies of relevant conditions can be explored by individual mode of acquisition or loss or between multiple modes. Figure 2 visualises the distribution of applicable conditions for the descent-based acquisition of citizenship by persons born in the country (A01a, left-hand plot) or abroad (A01b, right-hand plot).
The distribution of relevant conditions for descent-based acquisition of citizenship shown in Figure 2 demonstrates that in most countries, having a father or mother who is a citizen suffices to acquire the citizenship of that country: ius sanguinis is ‘generally applicable’ in 65 percent of countries if a child is born in the country and in 54 percent of countries if a child is born abroad (cf. Honohan and Rougier 2018, Figure 1 for a broadly comparable picture, with some differences due to different coding methodology). Besides gender- or wedlock-based restrictions, or group-based discrimination, countries also apply dual citizenship restrictions and/or what we call ‘sufficient link’ restrictions, where the acquisition of citizenship is dependent on the birthplace or place of residence of the child or parent.
Figure 3 visualises the distribution of the nominal years of residence required for residence-based acquisition of citizenship, based on a new variable A06a_yrs that records the precise number listed in the relevant legislation. We find that the most frequent nominal residence-requirement is five years (in 76 countries) and that most countries (163) have a residence requirement of a maximum of ten years. Fourteen countries have a higher residence requirement, up to thirty years in the United Arab Emirates or even forty years in Equatorial Guinea. In 15 countries, we did not identify a residence-based acquisition provision.
Fourth, the categorical coding scheme implemented in the Dataset also facilitates descriptive analysis beyond individual modes of acquisition by generating new categorical variables combining associated modes. We present two examples here.
Figure 4 visualises the distribution of countries for a territorial birth-based acquisition of citizenship variable, which combines information from modes A02a (ius soli for the ‘second generation’) and A02b (ius soli for the ‘third generation’, sometimes termed ‘double ius soli’). For simplicity’s sake, we leave aside additional related variables, such as acquisition by those who are born on a state’s territory and can acquire citizenship at a later age (A05). With that caveat, we see that unconditional ius soli, is most pronounced in the Americas and is otherwise rare, though still present in up to 19 percent of countries globally. Moreover, a form of conditional ius soli is present in another 17 percent of countries, where children born on the territory can only acquire the citizenship of that country, irrespective of the parental citizenship, if a parent was themselves born on the territory or fulfils a residence requirement. In twelve countries, gender-based or group restrictions apply.
In Figure 5, the distribution of countries is visualised based on a combination of one ‘acquisition’ and one ‘loss’ mode, which together reflect a country’s broad approach to dual citizenship acceptance upon naturalisation: mode A06b records whether a country requires applicants for naturalisation to renounce their previous citizenship (dual citizenship acceptance for ‘incoming’ naturalisation), while mode L05 records whether a citizen loses their citizenship if they voluntarily acquire another (dual citizenship acceptance for ‘outgoing’ naturalisation). In 2020, 60 percent of countries around the world fully accept dual citizenship, whereas only 24 percent consistently restrict dual citizenship at naturalisation. Almost a third of countries only accept dual citizenship for immigrants, i.e. incoming naturalisation (12 percent), or for emigrants, i.e. outgoing naturalisation (17 percent).
Where to go from here?
Having introduced the logic behind and possibilities of this newly systematised and updated GLOBALCIT Citizenship Law Dataset, it is important to emphasise that we see this as only version 1.0: we aim to continuously update the Dataset in order to provide the latest information on citizenship laws around the world. A next update is foreseen for 2022, to be worked out in collaboration with GLOBALCIT Country Experts and Regional Experts, for which we will update the data to 1 January 2022.
In addition, aside from these updates, we also plan – and have already started – to back-code the data in order to provide longitudinal information on citizenship law development. We will initially do so for selected indicators, such as birthright or dual citizenship (expanding e.g. this longitudinal dual citizenship dataset to cover not just automatic loss provisions, but also data on renunciation requirements for residence-based naturalisation). From there, our ambition is to gradually expand the longitudinal coverage to all relevant modes of acquisition and loss of citizenship.
Finally, and admittedly more tentatively, we aim to explore possibilities (e.g. building on the work of Thomas Huddleston) to cover not just legal provisions, but also policy practices that affect the acquisition or loss of citizenship. While such work cannot rely so directly on central coding and would require more extensive involvement of national and regional experts, we recognise that information on the practical implementation of citizenship law is vital for a truly comprehensive and contextualised understanding of the patterns, determinants, and consequences of citizenship regimes.
For a more elaborate introduction to the comparative methodology and presentation of some main patterns arising from the Dataset, see: Luuk van der Baaren and Maarten Vink, Modes of acquisition and loss of citizenship around the world: comparative typology and main patterns in 2020, GLOBALCIT Working Paper EUI RSC 2021-90.