The sociological study of formal citizenship has been strongly shaped by Brubaker’s (1992) analysis of citizenship legislation in France and Germany. Brubaker’s intuition was that citizenship laws express the conceptions of membership hold by governments and administrations, if not the society at large (although the author dissociated himself from the simplifications that subdivided nations in “ethnic” and “civic” – Brubaker, 1999).

Subsequent research on everyday citizenship, that is, the study of the conceptions of citizenship of the general population, has shown that the content of the laws does not necessarily match the larger ideas about membership in a society (see e.g. Miller-Idriss, 2006 on Germany and Sredanovic, 2014 on Italy). Further, the study of the bureaucracies of citizenship, while less developed, has shown that the bureaucracies sometimes hold conceptions of membership distinct from those in the law. Hajjat’s study of a French prefecture (2012) for example showed that the personnel examining citizenship applications saw the requirements of the law in terms of integration despite the letter of the law referring to assimilation. Further, the fact that an applicant wore a hijab, officially considered a proof of lack of assimilation both in court judgements and in ministerial circulars, was reinterpreted by the agents interviewed by Hajjat, either refusing to interpret whether the veil worn by the applicant was the “Islamic” hijab proof of lack of assimilation, or following other criteria than those indicated by the ministerial circular (see also Hajjat, 2010).

In this chapter I compare the conceptions of integration identifiable in the legislation of Belgium and the UK, and those expressed by the officers of different institutions working on the acquisition of nationalityFootnote 1 in the two countries. I will largely focus on “ordinary” acquisitions of nationality, that is, the provisions for those not born in the country and not benefitting from reduced requirements. A number of different provisions exist in the laws of the two countries, based on birth in the country, family links, age, disability, “exceptional” merits and, in the UK, on the complex colonial history (on the latter dimension see Sredanovic, 2017). Both Belgium and the UK have seen the introduction of restrictive requirements reflecting a large integrationist tendency in nationality policies across most of Western Europe (cf. Joppke & Morawska, 2003; Goodman, 2010; Rea et al., 2018). It is of interest to examine to what degree the officers working on nationality adhere to such integrationist approach, and whether the visions of integration of the agents are different from those expressed in the policies and in the political debate. Implementation research has shown that bureaucrats tend to apply their own notions of merit when applying policies (Lipsky, 2010). More specifically research on the implementation of migration policies, including residence procedures in France (Spire, 2008) and visa procedures in Belgian consulates (Infantino & Rea, 2012), have shown the tendency of bureaucrats to pursue what they perceive as policy objectives beyond, and even against, the letter of the law. In the cases I present here, the focus on integration in the political debate has left significant traces in the letter of the law, and has filtered down to a degree to the everyday activity of the officers I interviewed. However, the action of the officers seems to be much less oriented by notions of integration that one would expect: with a limited explicit mandate in the letter of the law to verify the integration of the candidates, the interviewees seemed to be more interested in verifying the formal requirements, and in other issues, such as fraud.

In the following pages I first offer some background on the general conditions and procedures for the obtention of nationality in Belgium and the UK, and details about the method of the research. I then discuss the several ways in which integrationist measures have found place in the nationality legislation of the two countries. Focusing on the interview data, I show the ways in which, despite significant space for discretion, the Home Office pursued in a limited ways ideas of integration (although Nationality Checking Services seemed more interested in the concept). I further show how officers working on nationality in Belgium did discuss notions of integration, but mostly within the limits of the formal requirements set in the letter of the law.

1 Policies of Naturalization

Here, I present the essential lines of the nationality legislation in Belgium and the UK, before analysing more in detail the integrationist measures present in both countries and the ways in which the concept of integration is managed by the officers implementing the law. In addition to the integrationist orientation present in both the countries, both Belgium and the UK had a two-step procedure at the time of the research, but the UK is an example of a high-discretion citizenship law, while Belgium is characterised by significant limits to the discretion of the officers examining nationality applications. While both the UK and Belgium have decentralised significant domains of policy, including, in Belgium, integration policy (Adam et al., 2018), nationality policy is exclusive domain of the central government in both countries.

In Belgium the nationality application normally requires five years of residence, and has to be submitted to the municipal register of the municipality of residence. The register theoretically should control the presence of all the documents required and transfer the application to the magistrates of the parquet – the local office of the procureur du Roi (“royal prosecutor”). In practice the municipal registers usually verified whether the applicant meets the requirements and often refused to transfer the application if they consider that some requirements are not met (cf. Sredanovic, 2020, 2022). Once the application reaches the local parquet the magistrates verify that it meets the requirements, which, given the “documentary” (Wautelet, 2014) approach of the Belgian law, in which almost all the requirements are expressed by the possession of specific documents, theoretically involves little work of appreciation. If the magistrates consider one of the requirements not to have been met, they can oppose the application, and every opposition can be appealed in the local court. Despite the limited discretion allowed by the law, the decentralised procedure brings to a significant local variation in interpreting aspects of the law such as the definition of employment and the degrees acceptable to fulfil integration requirements, as well as the infractions of the law taken into consideration (Sredanovic, 2020, 2022).

As mentioned, the UK is similar to Belgium in terms of presence of integration requirements and of a potential two-step procedure, but differs in the much larger discretion attributed to officers and in the centralisation of the procedure in a single entity located in Liverpool. In the UK an applicant has first to obtain indefinite leave to remain, which requires five years of residence and completing an English language test and a Life in the UK test. Once this status is obtained, all the applications for nationality are examined by the citizenship team of the UKVI (UK Visas and Integration, a division of the Home Office) in Liverpool. The applicant can choose to send the application directly to Liverpool, but could also choose to use a Nationality Checking Service, a service that local authorities could choose to activate to help with the applications. Starting with the end of 2018, however, the public services of the Nationality Checking Services were transferred to private operators. Differently from Belgium, the UKVI needs to give a formal acceptance of the application, as the procedure is discretionary, and only in limited cases a refusal can be appealed in court.

2 Methodology and Research Design

The ethnographic parts of this chapter are based on research conducted between 2016 and 2017 on the implementation of nationality legislation of the UK and Belgium. The main aim of the research was to analyse the discretion and the variation present in the implementation of the laws, but in this chapter I focus in particular on the ways in which integration was discussed (or not) in the interviews collected. The research project included in-depth qualitative interviews with officers working at different institutions linked with nationality acquisition. In Belgium this included 7 interviews with magistrates working in parquets and 23 interviews with civil registers across the national territory. In the UK I conducted interviews in 7 Nationality Checking Services in Northern England and Wales, and with 14 officers working on citizenship at UKVI in Liverpool (the latter interviews were conducted together with Émilien Fargues). I contacted most of the institutions involved in the research directly, aiming to represent the territory chosen for the two research projects (the whole national territory for Belgium, Northern England and Wales for the UK), although non-response from several institutions contacted means that, for example, Flemish institutions were under-represented.

The interview guidelines included questions about the organisation of work of each institution (number of employees, routines, volume of work, challenges in their work, formal and informal relations with other institutions), as well as about the interpretation of specific points of the law that emerged as potentially complex from the analysis of the laws and from the first interviews. The participants were asked to describe their own experiences. One important point is that I did not ask the interviewees about their personal opinions about nationality and immigration, as did for example Hajjat (2010) and Andreouli and colleagues (Andreouli & Stockdale, 2009; Andreouli & Dashtipour, 2014). The focus of the research was rather on the practices through which the law is implemented, and the visions analysed here emerged while discussing the everyday operations in applying the law. This means that the interviewees did not express all their opinions, but arguably only those they felt were both legitimate and relevant enough to be mentioned when discussing their work. While the personal opinions are certainly one of the factors that define how laws are implemented, other factors, including the organisation of work, routines, and institutional culture, can be equally or more important (see Sredanovic, 2020, 2022 for a more detailed analysis of the factors in the implementation of nationality law in Belgium and the UK).

The interviews have been conducted in English in the UK and in most institutions in Flanders, and in French in the rest of Belgium. Most interviews have been audio-recorded, always with the authorisation of the interviewees.Footnote 2

3 Integrationism in UK and Belgian Nationality Legislation

Citizenship legislation both in the UK and in Belgium has been influenced by the restrictive European tendency toward integrationism in citizenship laws that started in the late 1990s and accelerated after 9/11. Such a tendency requires the applicants for citizenship (but also for permanent residence and, in some cases, for family regroupment) to prove their cultural and linguistic integration, as well as participation in paid work (see e.g. Joppke & Morawska, 2003; Goodman, 2010; Rea et al., 2018). The wave of new requirements reached the UK early, with the New Labour 2002 Nationality, Immigration and Asylum Act introducing formal requirements for applicants to pass both a language and a “Life in the UK” test. Belgium was a relative latecomer, as it has introduced requirements of linguistic, economic and social integration only in 2012, and had a particularly inclusive law between 2000 and 2012 (de Jonghe & Doutrepont, 2013; Wautelet, 2014). Indeed, the 2012 reform has been identified as a shift from nationality as a tool of integration to integration as a requirement of nationality (de Jonghe & Doutrepont, 2013), with some Socialist and Green francophone politicians defending the former approach in the Parliamentary debates around the new law (Sredanovic, 2018). While there has been a clear tendency towards integrationism in Western Europe, cultural requirements existed even before (as noted also by van Oers, 2013): in Belgium cultural and linguistic integration, as well as lack of criminal past, were all examined before 2000 through police interviews. The UK policy was relatively less integrationist, but still formally required knowledge of English (or Welsh or Scottish Gaelic), which was verified only in exceptional cases and through interviews (van Oers, 2013). In other contexts, such as the Swiss one, naturalisation well before the integrationist wave involved procedures that were strongly based on conceptions of merit, and included highly invasive procedures such as circulating a profile of the applicant among all the citizens of the municipality (who often had the right to vote on the individual application), or in some cases the examination of the applicants’ high school report cards or bank accounts (Centlivres et al., 1991).

The language and “Life in the UK” tests in UK citizenship law have received extensive attention in literature, being recognised as a form of backlash against multiculturalism, as stigmatisation of the Muslim population in particular, after the 2001 riots in some Northern England cities and then the 2005 London bombings, and as the neoliberal promotion of an “active citizenship” closer to individual responsabilisation for the failures of the state than to a form of political agency (Kostakopoulou, 2010; van Oers, 2013; Turner, 2014; Puzzo, 2016; Byrne 2017). Not only the tests, but also the ceremonies and the oath have been recognised as part of this design (Byrne, 2012, 2014). A couple of observations could be added to this already substantial literature. While the reform of citizenship in the UK had a clear multicultural backlash content, two of the “autochthonous” minority languages, Welsh and Scottish Gaelic, were legitimated by their inclusion among the languages in which one can take the test (despite the latter having far less speakers in the UK than, for example, Polish or Punjabi). Secondly, while the New Labour reforms of citizenship were clearly inspired by a notion of cultural failure on the part of migrants and their descendants, the policy strived to express requirements in “civic” terms, insisting on learning “everyday” society skills. The “civic” approach of the New Labour government developed in a rather extremist direction with the 2008 Green Paper “The Path to Citizenship” which included plans for provisions such as probationary citizenship, increased residence requirements for those not engaged in volunteering, and barring applicants from obtaining nationality because of infractions committed by their children (Kostakopoulou, 2010). Such civic focus was reversed by the Conservative-Liberal Democratic coalition elected in 2010, which first abandoned the more radical civic plans of the Green Paper, and then explicitly culturalised the Life in the UK test in 2013 by rewriting the official manual to include more references to British history and culture (Turner, 2014; Puzzo, 2016; Byrne, 2017).

In Belgium, with a much higher level of multilingualism than the UK, one of the main issues was defining at what level should migrants integrate culturally. The current Belgian law requires explicitly linguistic and social integration, although the two are often linked. Both integration requirements are considered satisfied when the candidate has been in paid work during the qualifying period, recognising an integrative role of work (but see infra on contractual requirements). Linguistic integration can also be proved by passing a language test, while having obtained a high school or higher degree in one of the national languages in Belgium also fulfils both the social and the linguistic integration requirement. If the degree was obtained in one of the national languages and in an EU member state, it can be used to prove linguistic integration. The EU limitation is probably linked to issues of common educational frameworks, although it has the effect of othering, for example, francophone education outside Europe (and, presumably, in Switzerland). Both requirements can be further fulfilled by completing an integration course. As integration courses are defined by the three linguistic communities (Dutch-, French- and German-speaking, plus the Brussels Region, which is part both of the French-speaking and the Dutch-speaking communities, and therefore autonomous), this also means that migrants can integrate on a regional level. This was indeed part of the parliamentary debate about the new law. Most French-speaking parties (usually more centralist/unitary) were both sceptical about the integration requirements, and insisted on making the knowledge of any of the national languages the requirement. Most Dutch-speaking parties (usually more localist/autonomist) insisted that migrants should be required to learn the language of the region of residence (see Sredanovic, 2018). In the end the law requires the knowledge of one of the national languages, making the integration policy not entirely regionalist. One provision of the Belgian law opens theoretically to an integration evaluation on the part of the magistrates of the parquets. Candidates who have been resident at least 10 years in Belgium and prove the linguistic integration can have economic and social integration requirements waived if they prove the “participation to the life of the community”. Some examples of such “participation” are given in the ministerial circular (8 March 2013) to include formation or work in Belgium or the involvement in associations; the same circular categorically excludes associations linked to the country of origin, showing a fear of “separate societies”.

In addition to integrationist measures, the applicant is required to be free of dangerousness, a requirement that is extended in both countries by avoiding legislating in full on the matter. The UK has an extensive and inherently moralising requirement of good character (see Kapoor & Narkowicz, 2019 for an in-depth analysis). The January 2019 Home Office guidance includes among the signs of the lack of good character having been subjected to any sort of police measures, including fines and community sentences (which include court-ordered measures such as alcohol treatment and mental health treatment). Furthermore, such signs include irregularity in the migration history, as well as suspected criminal activity, associations with criminal or extremist organisations and individuals, bankruptcy, and having unpaid taxes or NHS charges. Failure on the part of the applicant to disclose one of these signs is constructed in itself as lack of good character. The officers are called to take in account the time elapsed since the event and disregard minor infractions, but also to identify patterns of “persistent” offences. Moreover, the officers are called to decide whether the sum of different signs suggest a lack of good character, which means that even aspects such as marital status, sexuality and lifestyle, which are defined as normally irrelevant, could be taken in consideration if they become “notorious”. In the Belgian legislation the concept of faits personnels graves (“serious personal infractions”) is used to define a number of factors that can bar from obtaining nationality. The (implementing) Royal decree names explicitly among such impediments for naturalisation crimes punished by a non-suspended prison sentence (as well as procedures that can bring to the same outcome). The decree further mentions explicitly having obtained one’s residence through fraud, while the law mentions condemnation for social and fiscal fraud. An explicit definition of crimes barring form naturalisation was a request advanced by the magistrates, who were struggling to implement the law (Apers, 2014). However, the courts’ orientation was that those given in legislation were only examples rather than a full list, leaving the situation open (see infra on the implementation). An even larger typology of dangerous behaviours included in the Royal decree derives from the reference to security issues as a basis for opposing a nationality application. The security issues are defined through a reference to the Law of 30 November 1998 on intelligence and security services. The latter, giving a large domain of activity to security services, mentions not only links with espionage, terrorism and criminal organisation, but also links with “extremist” organisations (including, among others, nationalist, “totalitarian” and anarchist organisations), and with “sectarian” organisations. It is clear that such a definition gives extensive discretionary powers regarding nationality policy to security services. While according to my fieldwork security-based oppositions do not seem to be largely used, there are individual cases in which applicants receive an opposition for issues of security, being unable to know the reason of such opposition (and therefore unable to challenge it effectively in court), always for security reasons (Wautelet, 2014).

Another point on which the existing literature on the “integrationist wave” has not focused is the degree to which States want their future citizens to be settled and relatively immobile. The Belgian law requires the qualifying period of five years to be covered by uninterrupted long-term (over three months) legal residence statuses, and limits the toleration of absences from the national territory to periods of six months maximum and/or a maximum one fifth of the residence requirement. The UK similarly tolerates a maximum of 450 days abroad in five years and a maximum of 90 days abroad in the last qualifying year, although some lenience can be discretionally exercised if longer absences are justified by the demands of one’s employer. Moreover, if the Belgian law is mostly satisfied with legal residence, in the UK, where legal residence is less clearly defined, candidates have to prove the physical presence on the territory, something that can be particularly challenging for candidates less subject to migration controls, such as (at least until the end of 2020) EU27 citizens. Moreover, the UK nationality law has an unusual explicit request that the new citizen makes the UK her or his main residence. While most of the migration literature in the last two decades has insisted on the capacity of migrants to form links with the country of residence despite continued links with the country of origin (e.g. Erdal & Oeppen, 2013) or despite high levels of mobility (e.g. Trenz & Triandafyllidou, 2017), states seem to continue to look at mobile people with suspicion, considering them apparently to lack authentic attachment to the future country of citizenship.

The UK law remains less integrationist than many in Europe by not including employment or income requirements – although such requirements are indirectly enforced through the requirements present in the immigration legislation, as well as through the exceptionally high citizenship fee (cf. Stadlmair, 2018 on how fees and economic requirements stratify by class the applicants). The Belgian law on the other hand explicitly requires most applicants to have spent most of the qualifying period in work, education or training – even profiles that in other countries are exempt from this kind of requirements, such as spouses of citizens, need to fulfil this requirement. While the formulation of the requirement recognises education and (some forms of) training along with work, it still excludes unpaid forms of work from definitions of integration. Further, while work is implicitly recognised as facilitating linguistic and social integration, only applicants that have worked uninterruptedly for the 5 qualifying years are exempted from proving separately linguistic and social integration. While Belgian governments, as most governments in Europe, have spent the last four decades reducing guarantees for workers and promoting temporary forms of work (see e.g. Bouquin, 2006), it is curious that in the nationality legislation candidates in regular, long-term employment are considered to be super-integrated.

4 UK: The Routinisation of Integration Requirements

In the the UK, the law does not require officers to verify integration, differently from what happens for example in France (see Hajjat, 2012). Instead, the UK law gives substantial discretionary powers to the Home Secretary: each naturalisation is a concession made by the Home Secretary, who can also decide to waive some requirements, including those relative to physical presence on the territory and absence of irregularities in the immigration history. Such large discretionary powers are largely exercised by the officers of the Home Office.

It is somehow surprising therefore that integration has a limited role in guiding the implementation of the nationality law within the Home Office. None of the case workers interviewed at the UKVI mentioned integration as something that was to be examined or more generally as an issue to be considered. As I discuss immediately below, much more attention was given to ordinary checking of the requirements and to the issue of fraud. Further, in an interview with two managers of the UKVI, direct questions about the role of integration in implementing the law were met with a certain perplexity, as examining the integration of applicants was not perceived to be a part of the mission of the citizenship team, and the integration aims were rather considered to be already fulfilled by the existence and the contents of the Life in the UK test. Upon hearing how “assimilation” was assessed in the procedures in France, the two interviewees further considered the procedure excessively subjective. Curiously, one of the few dimensions of integration mentioned by the two interviewees was that of the intention of remaining in the UK.

Considering the strong emphasis on integration behind the reforms, this could be read as a disjuncture between the policy and the implementation. However, such disjuncture is already present in the letter of the law, which indeed does not include indications to examine the integration of the candidates. The orientation of the officers of the UKVI is part of more general self-limiting approach in nationality questions, in which the ample discretion given by the law is reduced through the routinisation of decisions, aiming to have a relatively uniform treatment of the applications (Sredanovic, 2022).

The few exceptions in which some notions of merit, if not strictly of integration, did emerge, were linked to the discussion of “special” cases. Members of the armed forces, particularly if wounded in combat, were considered difficult to refuse nationality to – which meant that an application that fell short of some requirements was considered difficult to reject, although there were no special formal provisions existing for the naturalisation of the category.

A significant exception to the self-limitation in the use of discretionary powers existed in the examination of potential fraud. I have already underlined how the UK nationality policy is particularly stringent in regards to irregularities in the migratory experiences and to minor infractions. A similar stringency existed around the detection of possible fraud in the application. A number of caseworkers among those interviewed mentioned the importance of identifying fake documents among those submitted with the application – an issue that on the contrary was hardly ever mentioned in the interviews conducted in Belgium. Further issues of fraud were seen in reference to language proficiency. The results of the language test could be considered suspect if the test had been completed in a very short time and with few errors, and at the same time there were doubts about the actual competence of the applicant. However, communications from Nationality Checking Services about the lack of competence of an applicant – which were sent from time to time to Liverpool – could be disregarded if the local registrar spoke English with a strong local accent. The worries linked to suspicion of fraud in language test have been further exacerbated by a reportage of the BBC programme Panorama in 2014, in which some language test centres have been shown to allow sham successful tests in exchange for money. The reportage brought to a tightening of the management of the security of the tests, in order to make any kind of fraud harder (see Harding et al., 2020). This kind of worries had had an impact beyond the increased suspects of fraud, as a large number of foreign students have been expelled from the UK because the English language test necessary for their visa was taken in a language test centre considered to be a sham – although often there are no proofs, or indeed credibility, of an individual fraud (cf. York, 2018; National Audit Office, 2019).

I mentioned in the previous paragraph the abandonment of the 2008 Green Paper and of its radical civic approach to integrationism. The two managers mentioned above explained the abandonment of such plans by referring to doubts about requiring businessmen from other anglophone countries to volunteer in order to naturalise.

The integrationist orientation did however remain significantly inscribed in citizenship ceremonies, which do promote an image of the newly naturalised as integrated and participating to the local community (cf. Byrne, 2012, 2014). The workers of the Nationality Checking Services, who were often involved also in citizenship ceremonies, could therefore be closer to an integrationist approach than the officers working on nationality in the Home Office. An earlier research project conducted between 2007 and 2009, in which the importance of citizenship ceremonies was also noted, also included interviews with Nationality Checking Service workers (Andreouli & Stockdale, 2009; Andreouli & Dashtipour, 2014). In that research project, there was evidence that the officers of Nationality Checking Services expected the applicants to be employed and active in the local society (Andreouli & Dashtipour, 2014) – something that Andreouli and colleagues correctly observe was an official aim of the New Labour citizenship policies of the time, but, as I mentioned, is also something that was never included in the letter of the law. Some of the interviewees in the same research presented xenophobic positions that criticized the migrants as an economic burden or as reproducing separate cultures (Andreouli & Stockdale, 2009).

My interviews with Nationality Checking Service officers did not show a similar interest in integration issues, nor such integrationist attitudes. Such differences could be linked to the time passed between the two research projects, or even to geographical differences (London in the research of Andreouli and colleagues, Northern England and Wales in mine), but most probably the difference is linked to the interview approach. My interviewees arguably did not perceive that their personal opinions were relevant to their job. Indeed, such opinions could have brought some to decide whether to dissuade specific candidates, but Nationality Checking Service officers did not have a role as determinant as the officers of the Home Office.

When some ideas of merit or integration emerged, it was sometime in reference to marginal details, such as the quality of the filled-in forms

Officer 1: I mean, some people are so organised, it’s perfect, the form is so perfect, immaculate […] other people come and it’s a mess, it’s scribbled down…

Officer 2: Certain nationalities tend to be more organised, not always, but generally…

Officer 1: … yeah, there is a trend.

[British Nationality Checking Service 1]

Hajjat (2012) has observed how the neatness of the documents presented was a first detail according to which the officers working on nationality in his research in France started to categorize the applicants as integrated or not, but in UK case the impact of the judgement was significantly less. When judgments were made about the linguistic competence of the candidate – which, judging from the interviews conducted in Liverpool, brought some other local registrars to alert the Home Office – some interviewees underlined it was not up to them to evaluate this dimension:

Officer: … my colleague, she had an awful job, because [the candidate] hadn’t been able to complete the form, because his written English wasn’t very good, and his spoken English wasn’t very good. I don’t remember what nationality he was, I think he was Sudanese […]

Q: But he had passed the test…?

Officer: [nods] I don’t know how. That’s not for us to question, he had the certificates.

[British Nationality Checking Service 2]

This same officer, working in a Nationality Checking Service in Wales, on the other hand showed a particular appreciation of the choice of completing the application in Welsh, although such choice remained a hypothetical one, as she never saw an application of this kind.

The most explicit reference to deservingness emerged in relation to the regularity of the migratory history of the candidates.

Officer 1: … [migrants likely to have good character problems] mostly coming by lorry, not declared themselves at an official port, got caught by the police getting off a lorry on a motorway. That’s an illegal entry, whereas if they arrived at an airport and claimed asylum there, it could be slightly different.

[…]

Officer 1: … if you ask them [migrants with an unauthorised entry] the question, they don’t think they have done anything wrong, at all. Albeit they might have entered illegally, been here illegally. And they don’t see a lot of the time that that’s a problem. […]

Officer 2: [The applicant would say] “It’s normal”. “Does not everybody come on the back of a lorry”?

Officer 1: Oh we do hear that quite a lot, they [applicants] say that quite a lot.

Officer 2: “How else might I get in?”

Officer 1: If one applied for a visa, maybe.

[British Nationality Checking Service 3]

In the passage quoted the two officers were discussing the fact that unauthorised entries have been redefined as failing the good character requirement around 2014, and specifically the condition of many refugees, who, while obtaining a regular status thanks to a successful asylum procedure, can be still considered to have breached immigration norms through an unauthorised entry for the purpose of a citizenship application. These interviewees were quite knowledgeable of the complexity of different migratory experiences. However, this negative representation of refugees, and migrants more generally, who have entered the UK unauthorised, does not take in account how the visa regime excludes categorically a large proportion of potential migrants, including refugees, from authorised migration (see e.g. Neumayer, 2006).

5 Belgium: Integration as the Letter of the Law

In the Belgian context the “documentary” nature of the law (Wautelet, 2014) meant that the different institutions were hardly called to examine the degree of integration of the candidates. To a degree, by linking integration requirements to the possess of specific documents, the legislators have outsourced integration controls to external institutions (integration courses, educational institutions, etc.). The decentred procedure of nationality acquisition left space for significant local variations in the interpretation of the law (Sredanovic, 2020, 2022), but in the interviews I found a rather focus on the letter of the law, rather than on implicit policy objectives.

Visions of integration emerged sporadically in the interviews when discussing specific points of the law – the magistrates were more likely to interpret a specific point as linked to conceptions of integration, while the registrars were less likely to do so.

Some space for specific visions of integration was present when discussing the linguistic competence of the applicants. In one Flemish register, the officers interviewed justified the practice of the local parquet to oppose the nationality application of candidates who met the formal requirements but did not convince the police of their language competence during the interview.

Officer 1: [Some applicants] have the right documents but they don’t speak too good, to understand the interview with the police.

[…]

Officer 2: Because sometimes you have people who have been working for five years, and then that counts for work, it counts for language, and it counts for integration, but sometimes it’s people that just go through their work, are there all day, and then go to their home and don’t speak any Dutch, they can’t go to the supermarket and explain what they need, because they don’t speak the language. They get a letter from someone and they don’t know what to do with it. So, that’s why they say that you need to speak and understand at least a bit of Dutch to be able to interact with your community.

[Belgian register 1]

The practice is procedurally problematic, as it means that the explicit requirements present in the letter of the law are disregarded in favour of police interviews, and further there are no guarantees that the candidates lacking competence in Dutch do not speak French or German, which would also mean they meet the formal requirements of the law. In the hypothetical situation mentioned by the second officer of the register, the integration assumption is that such candidates might lack the basic notions of Dutch for everyday interactions, perhaps also because of long working hours that give little opportunities of learning the language. Nevertheless, even such minimal linguistic expectations bring the problem of why the assumptions of integration-through-work of the legislators should be disregarded, and who should examine the actual linguistic ability of the candidates, if the procedures dismiss the more qualified options of integration courses and language tests in favour of police interviews.

Some interviewees on the other hand criticized the norms introduced in the 2012 reform, as in the case of the interviewee of a Walloon local register (Belgian register 2) who considered the migrants in the territory to often be in an economically fragile situation, and resented having to examine the integration of people she already came to knew when they came to the municipality for other procedures (somehow implying that they were already socially integrated and did not have to prove it further). An interviewee in another Walloon register (Belgian register 3) highlighted the situation of those who took early retirement as an alternative measure to being laid off, and could neither prove economic integration nor access the lower requirements reserved for the over-65.

There are some differences along the linguistic divide in Belgium, as in Flanders there is a longer history of integrationist policies (Adam, 2013; Adam et al., 2018), as well as a stronger emphasis on the promotion of Dutch, which is perceived under attack from the increased use in French, especially in some municipalities around the Brussels region. However, this divide is linked only to limited differences, as ideas of integration more restrictive than those of the letter of the law emerged also in interviews with francophone registrars.

The magistrates showed a strong interpretative activity in the definition of faits personnels graves. As mentioned, the concept was codified following requests from the magistrates themselves, but the way in which the codification was expressed in legal sources was such that the definition was still considered open. In the parliamentary debate even the far-right Vlaams Belang did not advocate for anything less than a prison sentence to qualify as fait personnel grave (Apers, 2014; Sredanovic, 2018). Despite this, a number among the magistrates interviewed followed an extensive definition, including in some cases minor infractions such as traffic violations, and in one case police reports for which there was no decision to prosecute (see Sredanovic, 2020, 2022 for more details), following therefore a line closer to the more restrictive British approach. When deciding where to draw the line, also in relation to infractions that were distant in time, some interviewees seemed to draw from their work outside nationality. There have been critiques of the transfer of the migration portfolio to home affairs, often from ministries of foreign affairs or ministries of employment that formerly held the portfolio in many countries, the latter reflecting a once workforce-centric approach to migration (cf. e.g. Huysmans, 2000 for an EU perspective). The usual focus of home affairs on security and police arguably helps orient migration policies towards securitisation approaches that are less likely to recognise the rights of migrants. The fact that nationality (differently from the rest of the migration portfolio) is held in Belgium by the Ministry of Justice seems to give some guarantees to applicants. Still, the magistrates working on nationality also work as civil, and in some smaller parquets, as penal prosecutors. This seems to be reflected in some answers about the nationality applications that they consider opposing: often the considerations, e.g. whether the fine for the traffic infraction has been promptly paid, seems to follow the concept of “rehabilitation”.

Another case in which a number of magistrates included in the research (but not all) explicitly waive the law requirements is language requirements for nationals of confining countries (the Netherlands, France and Germany in particular), as, even when these candidates have no documentary proofs of the knowledge of the language, some magistrates are unwilling to oppose an application for that reason. Except in these cases, while some of the magistrates not included in the research did use police interviews to verify how candidates with documentary proofs speak the language, the magistrates I have interviewed followed more strictly the law. In the two following extracts candidates who do not have a full command of one of the Belgian national languages are partly problematised, but the legal norm, by which such competence has to be examined only documentarily, prevails:

A couple of weeks ago there was a person who according to his documents should be able to speak Dutch fluently, and who had brought his interpreter to court. Which made the president of the court laugh, but still he didn’t feel he could use this information to deny nationality to this person because it’s not in the law, if he has his attestation.

[Belgian magistrate 1]

I had a case of a Pakistani who works in a late-night shop and does not speak a word of French but works as self-employed since more than five years, and one can’t do anything. He is assumed to know the language.

[Belgian magistrate 2]

Despite doubts raised about the assumptions of integration inscribed in the law the magistrates interviewed prioritised the letter of the law over their visions of integration. It is also worth mentioning that Belgian magistrate 2 mentioned regretting the case of another applicant, unemployed but having worked in Belgium since the 1970s, and unable to pass through the 10-years route for a limited command of the national languages. More generally the 10-year route and the concept of “participation to the life of the community” it includes, while potentially a space for the officers to pursue their conceptions of integration, was largely considered a residual route both by the magistrates and by the registrars. As a consequence, very few kinds of proof of the participation in the life of the community were refused.

Similarly to the language cases of Belgian magistrates 1 and 2, in the interview with Belgian magistrate 3 there was a combination of personal interpretation of integration and priority given to the letter of the law. When discussing having to oppose the application of someone who had lived in Belgium for the qualifying period, but worked outside Belgium, the interviewee first said that “the law is the law” and then further argued the case in this way:

The meaning of the law is in any case to make one’s contribution to the Belgian economy. And so to work elsewhere it’s a bit complicated, even if it’s true that his life… I can understand the interpretation of [another institution] that said “listen, all his life is in the territory, and so all his money is spent here”, one could think he makes his contribution in this way, but it’s clear that the social security [contribution] is zero.

[Belgian magistrate 3]

While it is easy to consider the economic integration requirement as a duty to contribute to the national economy, in particular through taxation, there is little in the letter of the law that supports explicitly this interpretation. If anything, the economic integration concept seems to aim rather to avoid exclusion from the paid labour market.

Finally, while magistrates, as also the registrars, made mostly reference to the letter of the law when discussing the nationality procedures, there were some critiques of the general approach of the law:

Trying to put some things in legislation is really not easy; I mean, if you are talking about working, social integration, and such, what is acceptable? Because there are so many facets to a person, and if you come from a system [the pre-2012 one] where nationality was just granted as right, if you happened to meet the formal requirements of law, and you try to go back to more of a merit system… I mean, sometimes a person with all the right documents can have a lot less merit than a person who lacks some documents.

[Belgian magistrate 1]

This latter point in a way generalises the more specific critiques of the law advanced in Belgian register 1 and 2 above, pointing to the difficulty of translating complex conceptions of integration into law, and to the necessary mismatch between any conception of merit and the use of documentary proof. However, the question of whether the documentary approach is the real issue in defining access to nationality brings to larger question of implementation that I discuss in the conclusions.

6 Conclusions

I have shown how both in the UK and in Belgium the integrationist wave has brought to restrictive norms on nationality acquisition, that demand cultural conformity and lack of any sign of dangerousness from the candidates, that are hostile to highly-mobile lives and, at least in Belgium, privilege long-term forms of paid work. Previous research (Andreouli & Stockdale, 2009; Andreouli & Dashtipour, 2014) has shown some personal adhesion to such visions of integration among at least part of the personnel of Nationality Checking Services. Moreover, the present research has shown instances of following specific conceptions of integration in the implementation of the law, with the particular focus on fraud in the Home Office, and with the extension of the concept of faits personnels graves and the waiving of language requirements for candidates from neighbouring countries in some of the parquets in Belgium. However, despite the strong focus of the political and policy discourse on integration in both the countries, most of the interviewees in the research showed more interest in the letter of the law than in visions of integration. Most interviewees were likely to prioritise the letter of the law over the definitions of integration they themselves expressed, or to not consider their visions relevant at all to the description of their work. In Belgium such approach was reinforced by a nationality law that leaves limited space of discretion in its implementation. In the UK, the determinant factor seems an organisational preference for routinisation of decisions, despite a law that gives extensive discretional powers in assessing citizenship applications. As the Belgian magistrate 1 mentioned in the last paragraph put it, examining merit is difficult in general, and more so in approaches that attempt to follow uniform procedures. The main issue seems to stem from the will to introduce conceptions of integration, while the documentary approach taken in Belgium seems part of a trade-off inherent in any kind of implementation. Reducing discretion does necessarily make more difficult to waive the norm for “deserving” applicants who cannot match the requirements, but reduces also the space for bias and discrimination unavoidable every time the procedure calls for examining deservingness (a classic trade-off in street-level bureaucracy – see e.g. Lipsky, 2010).

The introduction of tests and formal integration requirements has introduced, both in the UK and in Belgium, new restrictions in the access of nationality that have put migrants at a disadvantage. One positive aspect of such policies is that at least they have codified integration requirements that in an earlier period and in other countries were already applied, sometimes in an even more invasive and arbitrary fashion (cf. Centlivres et al., 1991). Judging from the interviews collected for this research, there is some space in the implementation phase for transforming the integrationist ideology in more routinely applications of the letter of the law, although the mere fact of pursuing a charged and indefinite concept such that of integration introduces both procedural and substantial issues.