Nationality Policies in the Books and in Practice: Comparing Immigrant Naturalisation across Europe.

This article explores whether differences in the implementation of nationality laws explain gaps between nationality laws and nationality acquisition across Europe. Previous research confirms that nationality acquisition among the foreign-born in Western European countries is largely determined by the inclusiveness of nationality policies and the characteristics of the immigrant population. Drawing on recent data on nationality procedures across Europe, this article argues that, regardless of the nationality laws in place, nationality procedures create major obstacles for different types of immigrants. This article uses data from the European Social Survey to conduct multi-level analysis of nationality acquisition in 17 European countries among long-settled foreign-born adult arrivals, controlling for their individual, origin and destination country characteristics. The analysis finds that nationality procedures are as important as nationality laws for nationality acquisition among most immigrants living in Europe today, especially for immigrants who are most likely to benefit and apply: immigrants from developing countries.

Introduction

Comparative research has confirmed the importance of nationality policies and specific legal requirements for nationality acquisition among immigrants from developing countries (Janoski, 2010; Reichel, 2011; Gonzalez‐Ferrer and Morales, 2013; Vink et al., 2013). Nationality policies lead to a selection effect as they create certain opportunities and obstacles to acquire nationality. Immigrants’ reactions to nationality policies differ, depending on their abilities and motivations to acquire nationality. Nationality acquisition statistics suggest that it is not only abilities and motivations that determine nationality acquisition, but that immigrants with the same characteristics are more likely to acquire nationality in some countries with restrictive laws than in others with similar or even more inclusive laws.

In this context, nationality procedures remain understudied as a potential indicator of how nationality laws are working in practice. An implementation gap seems to exist in several European countries as the actual shares of nationality acquisition in various European countries are often higher or lower than expected, given the country’s legislation and immigrant population. Yet most comparative studies of nationality acquisition do not make a distinction between nationality laws and procedures (for an overview see Bilgili et al., 2015).

Figure 1 displays the correlation between the naturalisation laws, as measured by the Citizenship Law Indicators (CITLAW) in 2011, and the share of nationality acquisition among foreign‐born ESS respondents (2008–2016) who arrived 11–20 years ago from developing countries. This focus on residents with 11–20 years’ residence means that nearly all meet the de jure residence duration requirement for ordinary naturalisation.

Figure 1

Correlation between Nationality Laws and Acquisition among Foreign‐born Who Arrived 11–20 Years Ago as Adults from Developing Countries Source: CITLAW (2011), ESS wave 2008–2016

The correlation between the inclusiveness of nationality laws (as measured by CITLAW) and nationality acquisition shared among migrants from low/medium developed origin countries is positive but weak (0.11, p ≤ .05). Settled immigrants from developing countries seem more likely to acquire nationality in countries with restrictive laws than in others with similar or even more inclusive laws. At one end of the scale, countries above the expected trend line have higher than expected nationality acquisition shares based on their legal requirements: Denmark, Germany, Netherlands, Norway, Sweden and the United Kingdom. At the other end of the scale, below the expected trend line shows lower‐than‐expected shares of nationality acquisition in countries such as Cyprus, Greece, Ireland and Spain. How can countries with similar levels of inclusiveness or restrictiveness in their nationality policies end up with such different shares of nationality acquisition?

This article uses the formal differences in nationality procedures – from promotion to judicial review – as indicators of how much room the laws in the books leave for a potential implementation gap. Drawing on recent data on nationality procedures, the article argues that, regardless of the nationality laws in place, nationality procedures create major obstacles for different types of immigrants. This article uses a publicly available international dataset – the European Social Survey – to conduct multi‐level analysis in 17 European countries among long‐settled foreign‐born adult arrivals, controlling for their individual, origin and destination country factors. It uses the most recent ESS data (waves 2008–2017, 7472 foreign‐born who arrived as adults) and distinct measures of nationality laws and procedures: the Citizenship Law (CITLAW) and Implementation (CITIMP) indicators of the Global Citizenship Observatory (GLOBALCIT). The analyses find that nationality procedures are as important as nationality laws for most immigrants living in Europe today. Procedures are particularly important for vulnerable groups most likely to benefit and apply: immigrants from developing countries. We analyse the importance of nationality laws and procedures using multi‐level analysis looking at the immigrants group in our sample in general and for different sub‐groups in particular. This article’s conclusions argue that researchers and policymakers should give the same weight and attention to nationality procedures and changes in practice as they do to nationality laws and reforms on the books.

The remainder of the article is organized as follows. The next section reviews the literature on determinants of nationality acquisition, which illustrates the important gap regarding comparative research on the role of nationality laws and procedures. Our hypotheses are developed in the section after that.

Then data and methodology are discussed in detail, followed by a presentation and discussion of the results. Conclusions are provided in the final section.

Determinants of Nationality Acquisition

Immigrants’ origin and individual characteristics have been traditionally seen as the main drivers of nationality acquisition. Reviewing the international literature (see overviews in Chiswick and Miller, 2008; Liebig and Von Haaren, 2011), the three most significant characteristics are immigrants’ human and social capital, their chance to obtain dual nationality, and the level of human development of their origin country. All three shape nationality acquisition’s costs and benefits, which may be real or perceived, expected over the rest of their life or time in the country and relative to the benefits that they already have in their countries of origin and destination. For example, this cost‐benefit analysis is reshaped by the opportunity for dual nationality, which has boosted acquisition rates in the US and Europe (Yang, 1994; Jones‐Correa, 2001; Logan et al., 2012; Vink et al., 2013). Another important factor shaping this cost‐benefit analysis is the difference in the level of development of immigrants’ origin and destination countries – the greater the difference, the greater the likelihood of settling permanently and seeing greater benefits from nationality acquisition. Origin country development is so important that this factor conditions the relevance of most other individual‐level factors. In addition to these three main characteristics, new micro‐level studies are investigating the roles of family and social networks (Portes and Curtis, 1987; Logan et al., 2012; Street, 2014; Peters et al., 2016), political mobilisation (Jones‐Correa, 2001; Pantoja et al., 2001; Van Hook et al., 2006; Bloemraad, 2006a; Logan et al., 2012; Hansen and Clemens, 2018) as well as perceptions and experiences (e.g. Portes and Curtis, 1987; Zimmermann et al., 2009; Logan et al., 2012; Carrillo, 2015).

Nevertheless, these individual and origin factors are insufficient to explain differences in nationality acquisition across countries with similar immigrant populations, such as between Canada and the United States (Bloemraad, 2006b; Picot and Hou, 2011) or between European countries (Hagedorn, 2000; Rallu, 2009; Ersanilli and Koopmans, 2010). To explain this gap, most scholars point to differences in overall nationality policies, mostly in terms of the legal requirements on residence duration, dual nationality, economic, linguistic and civic integration. Differences in nationality policy have emerged as one of the most important factors in comparative European studies, both in terms of the annual acquisition rate and the overall share (Janoski, 2010; Sartori, 2011; Reichel, 2011; Gonzalez‐Ferrer and Morales, 2013; Vink et al., 2013; Stadlmair, 2017). The importance of nationality laws has also been confirmed in single‐country longitudinal studies of legal reforms, both in the US (Jones‐Correa, 2001; Balistreri and Van Hook, 2004; Van Hook et al., 2006; Bloemraad, 2006a; Fox and Bloemraad, 2015) and in Europe (e.g. Waldrauch, 2007; Anil, 2007; Van Oers, 2010; Green, 2012; Street, 2014).

The effect of nationality procedures has been largely understudied. Work by Bloemraad (2002, 2006a) first emphasized the potential role of these procedures. Bloemraad identifies major qualitative differences in American and Canadian practices. Her findings suggest that nationality procedures, in particular access to the procedure through promotion, information and documentation – could be as important an indicator of the institutional context for nationality acquisition. Single‐country studies have suggested that differences in procedures are also potential drivers of differences in nationality acquisition rates between different groups and local areas (e.g. Thränhardt, 2008; Logan et al., 2012; Hainmueller and Hangartner, 2013). In the absence of quantitative statistics, most researchers collect qualitative assessments through stakeholder interviews (e.g. Bloemraad, 2006b; Bauböck et al., 2013; Oliveira et al., 2017) – or, in rare cases, observation (Spire, 2008) and expert assessments (Chopin, 2006 and Huddleston, 2013). A few scholars have been able to quantify the importance of specific promotional measures and practical obstacles like fees (e.g. Felix et al., 2008, Hainmueller and Hangartner, 2013, Pastor et al., 2013 and Hainmueller et al., 2018).

Theoretical Framework

Nationality policies: the distinction between laws and procedures

Building on Bloemraad (2002; 2006b), nationality procedures are expected to create the context and lived reality for nationality acquisition. Nationality laws and nationality procedures respectively create de jure and de facto obstacles and opportunities for immigrants with specific characteristics. Whether or not immigrants are interested and able to apply and succeed may depend on the availability of information and courses, the costs, the types of documentation and assessments required as well as the amount of discretion, bureaucracy and judicial review. Within nationality procedures, the importance of “front‐end” practices regulating access may differ from the ‘back‐end’ difficulties in the procedure. Promotion measures and the documentation required for applications can be seen as the ‘front‐end’ of the bureaucracy that is most visible and impactful for immigrants. These symbolic and material resources “help with the practicalities of participatory citizenship and cement feelings of inclusion and attachment to the country” (Bloemraad, 2006b: p.237). Promotion and documentation affect all potential applicants. The ‘back‐end’ of the procedure, namely discretion, bureaucracy and judicial review, may affect the pace and predictability of the procedure once immigrants apply. These difficulties may lead to rejections and delays for certain applicants (see discussion of the interpretation of rejection rates in Bauböck and Helbling, 2011).

Hypothesis 1

The relative inclusiveness of both destination country nationality laws and procedures positively affects nationality acquisition among first generation immigrants.

The experience of vulnerable groups

Nationality laws and procedures are expected to have disproportionate effects on certain types of immigrants. The “citizenship premium” (see Chiswick and Miller, 2008; DeVoretz, 2008; Liebig and Von Haaren, 2011; Peters et al., 2018) refers to the return on investment experienced by certain types of immigrants whose acquisition of nationality leads to improved integration outcomes, usually measured in terms of their prospects on the labour market. Whether by nature or by design, nationality policies seem to create the greatest obstacles for the very immigrants most likely to apply for and benefit from this premium. Indeed, nationality policies sometimes explicitly refer to these immigrants as “vulnerable groups” who require more information and support, flexible documentation, exemptions and procedural protections. For example, refugees, who arguably enjoy the strongest relative benefits from nationality acquisition, are more likely to apply but are often dependent on government support to complete the process (Bloemraad, 2006b; DeVoretz, 2008). Nationality policies also mention groups like the stateless, disabled, elderly and young persons. In addition, quantitative analyses of nationality policies have identified the vulnerable groups mentioned below.

The main vulnerable target group is usually seen as immigrants from developing countries. The situation in an immigrants’ origin country emerges a major factor associated with the citizenship premium and with nationality policies. Vink et al. (2013) find that immigrants from developing countries are most likely to apply for nationality and enjoy this ‘citizenship premium’, but their acquisition rates are most affected by changes in nationality policy.

Hypothesis 2

The associations between inclusiveness of destination country nationality laws and procedures and nationality acquisition are positive for immigrants from developing countries, but not for immigrants from developed countries.

Research is less conclusive about the role that socio‐economic status plays for the citizenship premium and obstacles in nationality policies. Due to disproportionately high unemployment among the low‐educated and over‐qualification among high‐educated workers, both groups are over‐represented among low‐income immigrants who tend to benefit from wage growth after nationality acquisition. A clearer role for socio‐economic status would be expected for nationality policy. Individuals with a comparatively low socio‐economic status seem to be disproportionately excluded from costly, demanding and discretionary policies, according to evaluations from Austria, Switzerland and the United States (Hainmueller and Hangartner, 2013; Pastor et al., 2013; Reichel and Perchinig, 2015; Stadlmair, 2017; Hainmueller et al., 2018). Tertiary educated immigrants are expected to be better able to acquire nationality (see Yang, 1994; DeVoretz, 2008; Liebig and Von Haaren, 2011). Nationality laws and procedures often require, whether formally and informally, that applicants possess sufficient human capital to interact autonomously with state authorities and overcome any obstacles along the way without their support. Hence, socio‐economic status – measured in terms of education levels – is expected to create some legal and procedural obstacles, particularly for immigrants from developing countries.

Hypothesis 3

The positive associations between the inclusiveness of destination country nationality laws and procedures and nationality acquisition are stronger for low‐educated immigrants than for high‐educated immigrants.

Data and Methodology

Sample

In order to analyse the determinants of nationality acquisition, the multi‐level analysis in this article draws on the large‐scale international comparative European Social Survey (ESS).1 Our analysis includes a pooled data set from the five recent ESS waves of data collection (waves 4–8 data collection from 2008 to 2017).

For the purpose of this analysis, we define immigrants as foreign‐born, with two foreign‐born parents, adult arrivals aged 18 + in year of arrival, with at least six years of residence. Focusing on this group allows us to focus on the explicit decision to acquire nationality. We therefore exclude newcomers who are generally not yet eligible to acquire nationality as well as second generation migrants who may be able to acquire nationality by other means. Furthermore, we exclude child applicants, who are potentially able to apply through family‐based modes of acquisition, such as the extension of parental naturalisation. Similarly, six years was the closest mark possible with the ESS data to reflect the 5‐to‐6 year minimum and most common residence requirement for ordinary naturalization in the 17 countries sampled. Our final sample consists of 7472 individuals, including 6991 who could be matched to their specific country of origin. Specific sub‐groups were also selected based on the following independent variables: level of development of the origin country, highest level of education and cohort‐based residence duration.

The same country selection criteria were employed to construct a robust cross‐country pooled dataset: countries that participated in at least three of the five waves and sampled at least 80 immigrants according to this article’s definition. These criteria led to a dataset of 17 destination countries: Austria, Belgium, Cyprus, Denmark, Estonia, Finland, France, Germany, Greece, Ireland, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and the United Kingdom. Three additional Central and Southeast European countries (Croatia, Czech Republic and Slovenia) matched the statistical criteria for the overall sample, but we do not have information on the destination country nationality laws and procedures applicable to most immigrants who acquire nationality in these three countries. In these emerging destination countries, the share of nationality acquisition is reflective of their small immigrant populations mostly from neighbouring countries and their expansive co‐ethnic and extraterritorial nationality policies rather than their ordinary naturalisation policies under investigation in this analysis.

Dependent variable

The dependent variable is whether or not an immigrant has acquired destination country nationality among the population of interest. Nationality acquisition is measured at the time of the survey interview, while the laws and procedures are measured at one point in time. This dependent variable is well‐matched to ordinary and special naturalization policies, given the restriction of the target group based on the age at migration, residence duration, the place of birth of the respondent and their parents and exclusion of countries with primarily co‐ethnic‐based acquisition. The acquisition of nationality cannot be matched to specific ordinary versus special naturalization policies, because ESS does not capture the mode of acquisition or provide enough migrant‐specific variables for this matching, such as year of nationality application, nationality of spouse, refugee status or home language before migration.

Table 1 summarises the variation across the 17 European destination countries in terms of the share of nationality acquisition for immigrants from all countries, developing countries and developed countries. Most of Europe’s immigrants come from developing countries and they make up two‐thirds of the sample. The share of nationality acquisition varies considerably across the 17 European countries.2 The lowest shares are recorded in Greece (16.67% vs. 20.74% for immigrants from developed vs. developing countries) and the highest in Sweden (56.13% vs. 87.91%). The share of nationality acquisition among immigrants from developed and developing countries are strongly related. In other words, the shares of nationality acquisition tend to be higher or lower in the same countries.

Table 1

Share of Migrants Who have Acquired Destination Country Citizenship

	All		Low/Medium HDI	High HDI
	N	% with citizenship	% with citizenship	% with citizenship
Austria	217	37.19	42.74	30.86
Belgium	540	47.22	60.31	26.79
Cyprus	105	29.52	26.39	36.36
Denmark	225	44.44	50.69	33.33
Estonia	785	27.90	27.59	29.90
Finland	143	48.95	45.45	50.00
France	451	50.11	54.57	37.39
Germany	634	53.47	60.96	31.94
Greece	194	20.62	20.74	16.67
Ireland	801	27.97	25.55	26.91
Netherlands	379	67.28	79.44	32.35
Norway	195	44.62	66.04	19.10
Portugal	172	41.86	42.68	25.00
Spain	412	25.73	27.05	15.22
Sweden	535	72.71	87.91	56.13
Switzerland	1,098	39.16	38.04	40.10
United Kingdom	586	63.99	73.06	44.51
Total	7,472	44.16	47.73	35.93

Independent variables

This analysis’ independent variables include individual, macro‐level origin and destination country contextual variables. The same reference year is applied to macro‐level variables due to the nature of the dataset. For indicators capturing nationality policies, law and procedures, observations refer to 2011, which is the year closest to the midpoint of the five ESS waves and to the average year of interview.

The individual‐level variables are directly derived from the ESS. A few individual‐level variables merit further explanation. Educational attainment is measured according to the standard ISCED scale and categorised into low‐educated (elementary ISCED 1 and lower secondary ISCED 2), secondary (ISCED 3 and 4) and tertiary or high‐educated (ISCED 5 and 6). Residence duration is categorised into three dummies based on pre‐existing ESS coding of arrival cohorts: 6‐10 years, 11–20 years and more than 20 years. “Size of community” is a continuous variable calculated as the fraction of the total number of foreign‐born across all ESS waves from a specific origin country within the destination country. ‘Minority language speakers’ are defined as respondents whose self‐identified home language is not one of the official national or regional/minority languages in the destination country.

One additional individual‐level variable is included in order to assess the importance of laws and procedures for discriminated groups. The standard ESS questionnaire asks all respondents: “Would you describe yourself as being a member of a group that is discriminated against in this country?” For the purpose of this analysis we focus on the following aspects of discrimination: colour or race, religion and ethnic group.3

As for macro‐level variables, the relative socio‐economic status of the origin country is measured using the Human Development Index (HDI). This analysis uses the World Bank’s measures of low, medium and high developed countries and categorises them into developing (low or medium) and developed (high) countries. The analysis also controls for the relative socio‐economic status of the destination country in terms of GDP per capita in 2011 dollars. The variable ‘former territory or colony’ is retrieved from the public geo_cepii dataset.

The main policy variables of interest are the aggregate measure for nationality laws (CITLAW) and nationality procedures (CITIMP). The CITLAW score is coded as the average4 of the indicator for ordinary naturalization laws (ANATORD) and the indicator for special naturalization laws (ANATSPEC). ANATORD is a weighted average of six legal requirements: residence duration, renunciation, language, civics, criminal and economic. ANATSPEC is a simple average of law indicators for 16 types of acquisition. The CITIMP score is the standard ordinary naturalization aggregate coded by the GLOBALCIT national experts. CITIMP is the average of five dimensions of implementation that cover all stages of the administrative procedure in terms of promotion, documentation, discretion, bureaucracy and review (see Huddleston, 2013).

The main arguments for CITLAW and CITIMP are data availability and reliability. CITIMP indicators have not been calculated by GLOBALCIT for any of the 16 acquisition modes that make up the combined indicator on special naturalization laws. This analysis uses the ordinary naturalisation procedure as a proxy for all purposes for the purposes of cross‐country comparison. In addition to CITLAW and CITIMP, MIPEX ‘Access to Nationality’ (Huddleston et al., 2015) is used as a point of comparison based on an adjusted average, including only indicators relevant for the first generation (cf. Vink et al., 2013).

The dual nationality variable measures whether dual nationality is accepted by both the destination and origin country. This analysis combines data from the MACIMIDE Global Expatriate Dual Citizenship Dataset (Vink et al., 2015) for origin country policies and GLOBALCIT data for destination country policies.

The CITLAW and CITIMP indicators reveal a wide variation in nationality laws and procedures across Europe. The legal requirements for ordinary naturalization (ANATORD) include (1) economic and civic requirements in half of the European countries, (2) origin nationality renunciation requirements in a few, (3) a range of residence requirements from 5 to 10 years and (4) a range of language proficiency levels from none to B2. According to CITIMP, European countries have a wide variety of ordinary naturalization procedures, with a few countries with rights‐based procedures including no discretion and full judicial review, a few others with entirely discretionary procedures including no review and absolute discretion and most somewhere in between with few promotional measures but a wide range of documentation, assessments and bureaucracy.

Strikingly, countries differ widely in the relationship between their nationality laws and procedures. While a number of European countries have laws and procedures that are equally inclusive or restrictive, the countries with this disconnections between law and procedure seem just as numerous. Some countries show facilitated laws and procedures (Netherlands, Norway, Portugal and Sweden), some restrictive ordinary laws and procedures (Austria and Switzerland), some restrictive laws but facilitated procedures (Baltics and Germany) and others facilitated laws but restrictive procedures (Cyprus, Greece, Ireland, and Belgium).

The few reforms of naturalization laws and policies between 2007 and 2016 did not significantly change the score or ranging of the 17 countries, according to the CITLAW or MIPEX scales.5

Methodology

The analysis uses cross‐classified hierarchical linear models for a logistic multi‐level analysis that combines individual, origin and destination country variables, including nationality laws and procedures. This methodology is particularly appropriate as our sample is nested with respondents being linked to both a specific destination and an origin country. Cross‐classified hierarchical linear models capture the destination and origin country as parallel levels and correctly calculate standard errors at the contextual level (see Van Tubergen et al., 2004; Vink et al., 2013).

The results are presented in a stepwise‐approach in two tables, including 8 analyses for the 17 European destination countries in our sample. The four specifications in the first table display the results where the first specification includes the individual‐level variables, the second adds the origin country variables, and specification three and four introduce the policy variables with the MIPEX index in specification three and a differentiation between nationality laws and procedures in specification. The second table splits the sample in the different sub‐groups to investigate whether the effect of nationality laws and procedures differ by groups and test the group‐specific hypotheses. The first two specifications refer to the development level where the first specification is restricted to immigrants from low and medium HDI countries, and the second specification is restricted to immigrants from high HDI countries. The third and fourth specification explore to what extent the role of nationality law and procedures differ across educational levels, where the sample is split up in low educated, specification three, and high educated, specification four.

Results

Table 2 shows the results from the cross‐classified multi‐level analysis where the individual, contextual, and policy variables are stepwise included.6

Table 2

Logistic Cross‐classified Multi‐level Analysis of Destination Country Citizenship Acquisiton, in Odds Ratio

	(1)	(2)	(3)	(4)
Married (ref: single)	1.398***	1.440***	1.437***	1.439***
	(.115)	(.123)	(.122)	(.122)
Separated (ref: single)	1.079	1.095	1.093	1.094
	(.095)	(.101)	(.100)	(.101)
Gender (ref: male)	1.390***	1.414***	1.415***	1.416***
	(.082)	(.086)	(.086)	(.087)
Never children (ref: has child or children)	0.915	0.923	0.923	0.924
	(.059)	(.062)	(.062)	(.062)
Residence 6‐to‐10 years (ref: >20 years)	0.145***	0.150***	0.148***	0.149***
	(.016)	(.017)	(.017)	(.017)
Residence 11‐to‐20 years (ref: >20 years)	0.466***	0.467***	0.461***	0.466***
	(.040)	(.041)	(.041)	(.041)
Age	1.018***	1.020***	1.020***	1.020***
	(.003)	(.003)	(.003)	(.003)
Education (Lower Secondary)	1.018	0.961	0.965	0.961
	(.105)	(.103)	(.104)	(.103)
Education (Secondary)	1.170	1.153	1.162	1.157
	(.104)	(.108)	(.108)	(.108)
Education (Tertiary)	1.246*	1.275*	1.284***	1.281***
	(.112)	(.120)	(.121)	(.121)
Unemployed	0.921	0.987	0.983	0.986
	(.104)	(.116)	(.115)	(.115)
Community Size	0.967	0.907***	0.916***	0.907***
	(.022)	(.023)	(.023)	(.023)
Minority Language	0.484***	0.462***	0.469***	0.462***
	(.039)	(.039)	(.039)	(.039)
Discriminated group (ref: no)	1.111	1.120	1.122	1.124
	(.125)	(.132)	(.132)	(.133)
Origin country HDI	‐	0.962***	0.962***	0.962***
		(.006)	(.006)	(.006)
Former colony or territory	‐	1.782***	1.733***	1.777***
		(.209)	(.203)	(.208)
Destination country GDP per capita	‐	1.000**	1.000***	1.000***
		(.000)	(.000)	(.000)
Dual nat. policies (ref: not allowed)	‐	‐	1.052	1.093
			(.141)	(.151)
MIPEX Access to Nationality (1st gen)	‐	‐	1.023***	‐
			(.007)
Nationality Law	‐	‐	‐	1.020**
				(.008)
Nationality Procedure				1.018*
				(.009)
ESS fixed effects	Yes	Yes	Yes	Yes
No. of observations	7,472	6,991	6,991	6,991

p < .05;

p < .01;

p < .001; standard errors are in parentheses

The individual and origin country factors are robust across all specifications and the size and significance of the coefficients is in line with previous research. Immigrants who are currently married are more likely to be naturalised than those who are single or separated. Women are more likely to naturalise than men. The likelihood to naturalise increases with age and tertiary education. Immigrants are also less likely to acquire nationality if they come from larger origin communities in the destination country or do not speak the majority language. Having a child, being unemployed or belonging to a discriminated group is not associated with past nationality acquisition.

An important migration history variable is residence duration. The longer immigrants reside in the country of destination, the more likely they are to naturalise. Immigrants who reside in the destination country for 6 to 10 years are about 85 percent less likely to naturalise than immigrants who do so since more than 20 years. Immigrants who reside in the country of destination for 11 to 20 years are about 53 percent less likely to naturalise.

The origin country variables included in specifications (2) to (4) are significant. Immigrants from more developed countries are less likely to naturalise. An increase in 1 unit of HDI leads to a 3.8 percent decrease in the likelihood to naturalise. If immigrants’ origin country was a former colony or territory of the destination country, immigrants are more likely to naturalise.

Specifications (3) and (4) include the role of nationality policies. The introduction of these policy variables does not affect the strength or significance of the other individual or contextual variables. As previous research suggests, the country’s overall nationality policies, measured by MIPEX, is a highly significant predictor of nationality acquisition. A 1‐unit increase in a country’s policy on the MIPEX 100‐point scale on “Access to nationality” is associated with a 2.3 percent increase in nationality acquisition. Replacing MIPEX with measures of nationality law (CITLAW) and nationality procedures (CITIMP) demonstrates that nationality laws and procedures matter independently of each other. A 1 unit increase in a country’s law on the CITLAW 0‐1 scale is associated with a 2 per cent increase in nationality acquisition. The same increase in the CITIMP 0‐1 scale is associated with a 1.8 per cent increase. Thus, if the nationality laws are very inclusive, the likelihood to acquire nationality increases by two per cent. If procedures are very inclusive, the likelihood to acquire nationality increases by 1.8 per cent. This finding confirms our hypothesis H1 that nationality laws and procedures have an independent, but similarly strong and positive effect on nationality acquisition. The opportunity to have dual citizenship, based on the origin country policies, does not affect the likelihood of nationality acquisition in both specifications (3) and (4).

Table 3 represents the results of the cross‐classified multi‐level models for different sub‐groups to investigate whether the effect of nationality laws and procedures differ by groups and test the group‐specific hypotheses. The different groups are low/medium and high HDI countries (specification 1 and 2) and low and high‐educated immigrants (specification 3 and 4).

Table 3

Logistic Cross‐classified Multi‐level Analysis of Destination Country Citizenship Acquisiton by Sub‐groups, in Odds Ratio

	(1)	(2)	(3)	(4)
	Low/Med HDI	High HDI	Low education	High education
Nationality Law	1.035***	1.002	1.024**	1.014*
	(.011)	(.006)	(.009)	(.006)
Nationality Procedure	1.036***	1.001	1.031**	1.021**
	(.009)	(.007)	(.013)	(.007)
Dual nationality (ref: not allowed)	0.935	1.705**	1.087	1.558*
	(.163)	(.372)	(.243)	(.278)
ESS fixed effects	Yes	Yes	Yes	Yes
No. of observations	4,680	2,311	1,941	2,444

All specifications control for marital status, gender, children, years since migration, age, educational level, unemployment, community size, minority language, discriminated group, origin country HDI, former colony, destination country GDP, ESS waves.

p < .05;

p < .01;

p < .001; standard errors are in parentheses

The individual and contextual variables are the same compared with Table 2, except that the variables marriage and tertiary education only remain significant for immigrants from developing countries. Comparing Specification (1) and (2), destination country’s laws and procedures seem to matter greatly for immigrants from developing countries, but little to immigrants from developed countries. Among immigrants from countries with a low or medium HDI nationality laws and procedures are of equal importance. If the nationality law index increases by 1 unit, the likelihood of nationality acquisition increases by 3.5 per cent, the same increase in the nationality procedure index, increases the likelihood by 3.6 per cent. This positive and similarly strong effect of nationality law and procedure cannot be observed for immigrants from developing countries. This supports Hypothesis H2 that the effect of nationality laws and procedures is stronger for immigrants from developing countries as they might have a stronger interest to secure their legal status in the destination country than immigrants from developing countries. Dual nationality policies in the origin country, by contrast, are the only policies that matter for immigrants from developing countries.

Comparing specification (3) and (4) shows that favourable nationality laws and procedures positively affect the likelihood to naturalise for low‐ as well as for highly educated immigrants. However, the nationality laws and procedures both matter more for low‐ and highly educated which confirms Hypothesis H3. A 1‐unit increase in the nationality law index increases the likelihood to naturalise by 2.4 percent for low‐educated and 1.4 for highly educated. If the nationality procedure index increases by 1 unit, the likelihood to naturalise for low‐educated increases by 3.1 percent and for high‐educated by 2.1 percent. The differences in the effect between highly and low educated is illustrated in Figure 2 which shows the predicted probability of citizenship acquisition depending on naturalisation laws and procedures. The line of low‐educated is steeper, even more pronounced for nationality procedure, than the line for highly educated, which indicates the overall stronger effect for low‐ than highly educated migrants. This finding suggests that higher educational levels decrease the role of legal and procedural obstacles.

Figure 2

Predicted Probability of Destination Country Citizenship

The results presented so far have shown that nationality laws and procedures affect different groups of immigrants to different extent. CITIMP is differentiated in terms of access to procedures (i.e. promotion and documentation) and the difficulty of procedures (i.e. discretion, bureaucracy, and review) in order to explore which of those two dimensions drives the effect of nationality procedures.

As the results in Table 4 show, practices regulating access to nationality procedures seem to drive the effect of nationality procedures for the more vulnerable groups – immigrants from low/medium HDI origin countries and immigrants that are low educated. For highly educated immigrants, on the other hand, the role of procedural difficulties seems to be more pronounced. The results were re‐run by residence duration cohort in order to assess the significance of laws vs. procedures. Laws seem to have an exclusionary function, regardless of their residence duration. In contrast, procedures seem to have a delaying function only for immigrants with 11–20 years. This cohort with 11–20 years of residence corresponds to recently eligible immigrants based on the maximum 10‐year residence requirement.

Table 4

Logistic Cross‐classified Multi‐level Analysis of Destination Country Citizenship Acquisition Including Access and Difficulties of Nationality Procedures, in Odds Ratio

	(1)	(2)	(3)	(4)	(5)
	All	Low/Med HDI	High HDI	Low education	High education
Dual Nationality	1.096	1.005	1.880*	1.342	1.558*
	(.151)	(.164)	(.403)	(.356)	(.279)
Nationality Law	1.020**	1.035***	1.002	1.031***	1.014*
	(.008)	(.009)	(.006)	(.009)	(.006)
Nationality Procedure ‐ Access	1.015	1.029*	1.004	1.050***	1.009
	(.010)	(.011)	(.008)	(.011)	(.007)
Nationality Procedure ‐ Difficulty	1.005	1.012	.998	1.007	1.012*
	(.009)	(.010)	(.007)	(.010)	(.006)
ESS fixed effects	Yes	Yes	Yes	Yes	Yes
No. of observations	6,991	4,803	2,397	896	2,444

All specifications control for marital status, gender, children, years since migration, age, educational level, unemployment, community size, minority language, discriminated group, origin country HDI, former colony, destination country GDP, ESS waves.

p < .05;

p < .01;

p < .001, standard errors in parentheses

Conclusion and Discussion

In order to explain nationality acquisition among immigrants in Europe, countries’ under‐appreciated nationality procedures are as important as the well‐studied nationality laws. This initial exploration of 17 European countries’ laws and procedures as of 2011 suggests that nationality laws and procedures are two very important factors driving nationality acquisition in Europe, where nationality procedures emerge as a distinct factor that is complementary to nationality laws and equally strong for all immigrant groups. De facto obstacles in procedures may be as important as de jure obstacles and unrelated to the laws in the books. In general, the likelihood to acquire nationality is (1) lower under restrictive laws and procedures (e.g. Austria and Switzerland), (2) highest under inclusive laws and procedures (e.g. Netherlands, Portugal, and Sweden), (3) somewhere in between when restrictions exist in either the law (e.g. Estonia and Germany) or procedure (e.g. Cyprus, Greece and Ireland).

Like nationality laws, nationality procedures affect those immigrants that most likely apply and benefit: immigrants from developing countries. Nationality procedures matter as much as nationality laws for immigrants from developing countries, based on the 2011 laws and procedures in 17 European countries. As expected in hypothesis 2, most laws and procedures tend not to matter for immigrants from developed countries who acquire nationality. The one exception is dual nationality. Dual nationality may be seen as a necessary precondition for nationality acquisition among immigrants from developed countries and tertiary educated immigrants from developing countries, both of whom have greater opportunities for successful return plans and comparable standards of living and well‐being. Contrary to hypothesis 3, both nationality laws and procedures matter for low‐, secondary‐ and tertiary educated. Laws and procedures can have an exclusionary or delaying function to discourage or delay immigrants from developing countries, despite their interests in the citizenship premium. The finding that procedures are most relevant for recently eligible immigrants indicates that even immigrants who can meet the legal requirements face additional obstacles to prove that they meet them in practice.

This research has confirmed the international relevance of Bloemraad’s seminal work studying nationality procedures and procedures and the few single‐country case studies of procedures. This article shows how laws and procedures both affect the naturalisation decision of migrants. We hope that with future data sets of a greater sample size the relation between laws and procedures can be further explored. Existing studies have mostly focused on access to procedures, such as the fees and exemptions (Felix et al., 2008; Pastor et al., 2013; Hainmueller et al., 2018), rather than the difficulty of procedures themselves in terms of decision‐making structures and judicial review (Hainmueller and Hangartner, 2013). Future research on nationality acquisition should therefore expand to this understudied field, particularly into the relevant importance of procedural access vs. difficulty for different immigrant groups. Respondents need to be matched to their eligible law and procedure, which requires their specific year of arrival, year of eligibility for naturalisation and CITIMP coding of all major modes of acquisition. Future research should also better capture vulnerable groups, such as discriminated groups, low‐income and refugees, which the few single‐country studies suggest are most sensitive to obstacles in procedures.

Policy actors should also give greater attention to nationality procedures and practices. In most European countries, ordinary naturalisation usually involves considerable bureaucracy and documentation from the origin and destination countries, with few legal exemptions for vulnerable groups (Huddleston, 2013). Improvements in these practices, often overlooked in policy and public debates, may be precisely what thousands of eligible immigrants need to become citizens and contribute to the future of Europe’s diverse democracies.