Contents

Chapter 1. Research Purpose and Methodology

Tineke strik1.1Purpose of the Study1.2Methodology1.3Interviews

Chapter 2. Integration Tests Abroad

Tineke Strik2.1Organisation and Costs2.2Level and Content of the Test2.3Target Groups and Exemptions2.4Purposes of the Tests2.5Consequences of not passing the test2.6Effects of the Tests: Statistical Data2.7Effects of the Tests: Interview Results2.8Summary and Conclusions

Chapter 3. Integration Tests in the Country

Anita Böcker3.1Description of the Requirements3.2Preparation for the Examinations3.3Costs3.4Target Groups and Exemptions3.5Consequences of not passing the test3.6Purposes and Debates3.7Effects of the Requirements: Statistical Data3.8Effects of the Requirements: Interview Results3.9Summary and Conclusions

Chapter 4. Integration Tests in the Naturalisation Procedure

Maaike Luiten and Ricky van Oers4.1Description of the Tests4.2Preparation for the Tests and Costs4.3Exemptions4.3.1Exemptions on the Basis of Sufficient Integration4.3.2Exemptions on the Basis of age or an impediment4.4Purpose of the Tests4.5Consequences of not passing the Test4.6Effects of the Test: Statistical Data4.7Effects of the Tests: Interview Results4.8Summary and Conclusions3

5

578

11

1116172129303338

45

455152535557616672

77

778286868890939399105

This report presents a comparative study of the policies, practice and data regardingthe integration requirements of nine EU Member States. Language and integrationtests as a condition for naturalisation and various types of legal residence permits aretopical issues in several EU Member States. A number of common trends regardingthe integration of immigrants from third-countries are visible.Firstly, several Member States have introduced compulsory integration coursesand tests as a condition for admission for family reunification, permanent residence,or naturalisation. Secondly, in certain Member States there appears to be a trend to-wards formalisation of the language and integration requirements. The informal in-terview between an applicant and a civil servant as part of the naturalisation proce-dure has been replaced by a formalised test, which often coincides with a raising ofthe level of required knowledge of the language and of the host society. Thirdly,there appears to be a readiness among Member States to learn from the experiencesof other Member States and to copy measures developed elsewhere in the EU. Theexchange of information in the Justice and Home Affairs Council, the adoption of theCommon Basic Principles on Integration in 2004 and of the Commission’s CommonAgenda for Integration in 2005, and the informal meetings of EU Ministers responsi-ble for integration have contributed to the openness of experiences of other MemberStates. Furthermore, most Member States have recently been confronted with thequestion whether to introduce integration measures or integration conditions in theprocess of implementation of Directive 2003/86/EC on the right to family reunifica-tion and Directive 2003/109/EC on the status of long-term resident third-country na-tionals. Both directives make reference to the possibility of introducing integrationmeasures or requiring immigrants to fulfil an integration condition in order to be en-titled to a certain status provided for in those directives. The introduction of the testsreflects a change in ideas on the relationship between legal status and integration.Although the imitative behaviour of the Member States with regard to integra-tion requirements is rapidly increasing, little is known of the actual effects of this ra-ther recent legislation. Mostly the national legislator mentions the integration of im-migrants in the host society as the main purpose of these requirements. Whether theintroduction of these new ways of testing integration has actually contributed to bet-ter immigrant integration or whether the tests function as a mechanism for selectionand exclusion is unknown. Exclusion results in migrants being denied admission to acertain Member State and not obtaining permanent residence status or naturalisation.It is also unclear to what extent being put through an integration test affects the be-haviour of potential applicants or how such a requirement is perceived by them. Sofar there have been few if any empirical studies on the actual effect and impact of theuse of integration and naturalisation tests.5

The INTEC project has conducted a first evaluation of the recent paradigmaticchange in policy concepts concerning integration in certain Member States. Theproject aimed to provide detailed and reliable information on the content and theimpact of compulsory elements in national integration policies. The main researchquestions focused on the reasons for the introduction of the obligatory integrationrequirements, the way in which they had been developed and put into practice andthe actual effects of the requirements on immigrants. Furthermore attention was paidto the differences between integration requirements with a voluntary and a compul-sory nature, the relationship between the different requirements in a Member Stateand the relationship between the integration requirements of the different MemberStates. Within the framework of the INTEC project, research was conducted on thenational policy and practice concerning these tests and their effects in nine MemberStates: Austria, Belgium, Denmark, France, Germany, Hungary, Latvia, the Nether-lands and the United Kingdom, Hungary and France.1The rapporteurs of the twolatter Member States conducted the research at their own expense, which forcedthem to limit the research to a more moderate scale.2On the basis of these nationalreports, the Centre for Migration Law made a comparative study, which is laid out inthis report. As this study has analysed the impact of compulsory elements in nationalintegration policies, it allows judgement of the contribution of those tests to the inte-gration of the persons concerned and the possible intended and unintended effects ofthe introduction of the tests.The central aims of the INTEC project have been to collect, analyse and dissemi-nate accurate, systematic, and up-to-date information on the practices and effects ofthe integration and naturalisation tests in Member States that have introduced suchtests. The outcome of this research enables public authorities, both at the nationaland local level, politicians, immigrants and their organisations as well as academics,journalists and members of the public to form an informed opinion on the differentkinds of language and integration tests introduced in the Member States, their con-tribution to the integration of immigrants, the possible effects of the introduction ofthese tests and the arguments pro and contra such tests. In this sense the INTECproject may increase the Member States’ capacity to develop, implement, monitorand evaluate policies and measures for the integration of third-country nationals andstimulate transnational exchange of good and bad practices with regard to the inte-gration of these nationals. Furthermore, this project aims to disseminate the results ofnew policies of Member States, not only between the nine Member States concernedbut also to all Member States. In order to ensure the dissemination of the research re-

As Denmark does not take part in the EU Justice and Home Affairs acquis, the Integration Fundwas not able to finance the research conducted in this Member State. To enable the involvementof the experiences in Denmark, the Danish Institute for Human Rights has conducted half of theresearch at its own expense. The Dutch Centre for Migration Law has financed the other half ofthe expenses for the Danish research.The French rapporteur was not able to conduct interviews and the Hungarian rapporteur used alimited number of interviews (but instead,of interviewing immigrants she assessed 100 files re-garding applications for permanent residence or citizenship). Both rapporteurs will not organisea national seminar on the outcome of their research.

The basis of this comparative research is formed by the national reports on nineMember States. These national reports, which encompass more detailed information,are therefore a substantial part of the INTEC project. The authors of the reports arealso responsible for the field research in their Member State.3The researchers alreadyhad extensive research experience on this topic before the INTEC project started.Most of them had published widely on issues of integration, immigration or national-ity law in their home country, participated in comparative studies or had themselvesconducted comparative studies on integration programmes covering several MemberStates. Four of the institutions involved participated in the NATAC4comparativestudy on the nationality law and practice in 15 EU Member States (Bauböck, Ersbøll,Groenendijk and Waldrauch 2006). Most of the rapporteurs had also contributed to astudy on the national policies concerning the integration of newcomers and/or futurecitizens in their countries (Van Oers, Ersbøll and Kostakopoulou 2010). These rele-vant experiences have contributed to the completeness and quality of the nationalreports and the synthesis report.The national rapporteurs started their cooperation in a public kick-off seminar inFebruary 2010, where experts on integration policy, linguists, sociologists and civilservants gave an overview of conducted research and formulated recommendationsfor the INTEC project. In a closed setting, the rapporteurs deliberated on the ques-tionnaires for the field research, the selection of the respondents and the criteria andformat for the national reports.In order to formulate the aims of the integration and naturalisation tests, therapporteurs analysed the decision making process regarding the integration and na-turalisation tests. They included in this analysis the political debates on the actualbills and comments from advisory bodies, immigrant organisations, academics andother experts. The analysis offered an overview of the arguments and foundations ofthe introduction of the tests and the way criticism or possible risks had been taken in-to account. Furthermore the content of the tests, the practical implementation, thetarget groups and exemptions were described. Using a number of different sources,the rapporteurs investigated the impact of the tests. They assessed criticism and rec-ommendations of (international) experts in literature and reports that appeared afterthe entry into force of the relevant acts and analysed the jurisprudence on the inte-34The national reports are separately available on the website of the Centre for Migration Law,www.ru.nl/rechten/cmr. The national rapporteurs are mentioned on page 124.EU funded research in 2004-2005 on the acquisition of nationality in EU Member States: rules,practices and quantitative developments.

gration requirements. Furthermore evaluations of the tests and other related studiesand data on the effects of the tests were assessed, including the political reaction tothe outcome of the evaluations. The data and number of studies differ per countrybut also depend on when the tests were introduced. As the naturalisation tests wereintroduced first, data and literature on this topic were relatively widely available. Insome Member States the availability of evaluations was limited or even absent be-cause of the recent introduction of the tests, which was especially the case with theintegration and language tests abroad. This was one of the reasons why the rappor-teurs also conducted field research in order to collect information on the effects of thetests. Furthermore, the material collected by the interviews offered clarification of thefigures, as they revealed the perception by the migrants and other respondents of theintegration requirements.Data collection for the national report as well as this synthesis report ended inOctober 2010.

In eight countries, a total of 329 interviews were conducted. Most of the interviews(about 25 in each country) were with immigrants who had been or were required tofulfil integration requirements. In addition, interviews were conducted with teachersof integration courses (about 5 per country); public officials responsible for the appli-cation of the integration requirements or the naturalisation legislation (about 5 percountry); and staff or active members of immigrant organisations and other NGOs(also about 5 per country). The Dutch research team conducted additional interviewsin Turkey, with candidates of the Dutch integration test abroad (10) and their teach-ers (3). Table 1.1 gives an overview of the interviews conducted in each of the eightcountries.In most countries, the interviews with immigrants were arranged through differ-ent channels, and the researchers strived for diversity among the respondents interms of, e.g., gender, nationality, age and educational level. However, as it was notpossible to use interpreters, the research teams in most countries could only conductinterviews with immigrants who were sufficiently proficient in the language of thecountry of immigration or, e.g., English. Table 1.2 provides information about themethods used for contacting potential respondents and the characteristics of the im-migrants interviewed in each country.Some of the national research teams used additional data collection methods:- The Hungarian team analysed 100 files of naturalisation applications in the Of-fice for Immigration and Nationality Affairs.- The UK research team made a freedom of information request to the UK BorderAgency to obtain statistical material concerning the ‘Life in the UK’ requirement:(1) pass rates by nationality; (2) the share of persons obtaining indefinite leavewho rely upon the test, ESOL study and an exemption; and (3) recent trends inthe number of persons obtaining indefinite leave and naturalisation.- The Danish research team made a request to the Danish Ministry of Integrationto obtain statistical data concerning granted and rejected applications for perma-8

Table 1.2 Interviews with immigrantsNumber ofinterviews25Arranged through:language centres andimmigrant advice cen-tres in Vienna andWiener NeustadtRespondent characteristics16 women, 9 men; 13 different national-ities; mostly admitted for family reuni-fication; half were aged 25 or younger;half had post-secondary education;most had lived in Austria for about ayear12 living in Antwerp (obliged to attendintegration course), 8 in Brussels (noobligation); 13 women, 7 men; between26 and 49 years old; 14 different natio-nalities- 12 applicants for naturalisation:9 women, 3 men; nine different natio-nalities; half were born or raised inDenmark; half were aged 25 or young-er; 10 with a middle or high education-al level, 2 with a low educational level;- 14 applicants for permanent resi-dence:9 women, 5 men; ten different national-ities; most had come to Denmark forfamily reunification; all but one wereaged above 25; 11 with a middle orhigh educational level, 3 with a loweducational level.

CHAPTER2: TESTABROADNumber ofinterviews27Arranged through:adult education cen-tres and test centresRespondent characteristics8 had passed the integration test abroadas a condition for admission; 12 hadpassed the integration test in Germany;10 had passed the integration test fornaturalisation;12 women, 15 men; 18 different natio-nalities; between 22 and 50 years old, 3younger than 185 had passed the naturalisation exam;the others were taking it or were at-tending a course to prepare for it15 naturalised citizens, 13 non-citizens25 were taking the integration exam(not always clear whether for naturali-sation, permanent residence or obliga-tory, the test is the same); 1 had juststarted a course, 1 just had applied fornaturalisation, 1 was exempted, one didnot apply for a permanent residencepermit or naturalisation were ; 15 dif-ferent nationalities (13 were Turkish);12 between 20 and 30 years old, 12 be-tween 30 and 40 years old; 20 women, 9men.(7 had already passed the integrationtest abroad in Turkey; 3 were still at-tending a course to prepare for it) (4 be-tween 20 and 30 years old, 4 between30 and 40 years old) (6 women, 4 men)(5 had academic or higher vocationaleducation)all 16 had taken the ‘Life in the UK’test;8 women, 8 men; 11 different nationali-ties; between 21 and 61 years old

Chapter 2. Test AbroadTineke StrikMore and more Member States tend to require a certain knowledge of their languageand/or their society as a condition for admission. In 1990, Germany was the firstMember State to apply a language requirement for admission, although it was li-mited to children between 16 and 18 years who applied for subsequent migration inorder to reunite with their parents in Germany. In 2005, the German governmentstarted to subject two preferential groups to a language test: the spouses and descen-dants of the so-calledAussiedler(ethnic German applicants) and Jewish immigrants(as well as their spouses and descendants) who were aged 15 or older before entry.After having passed the test, the Aussiedler and their family members receive Ger-man nationality and the Jewish immigrant receives a permanent residence permit. In2006, the United Kingdom started to require language skills from highly skilledworkers and the Netherlands introduced a language and integration test for appli-cants for family reunification. This last example has been followed and is going to befollowed by more and more Member States.At the time of writing of the report, a language requirement for family reunifica-tion is applied in the Netherlands and Germany. Furthermore, the Netherlands re-quires a certain knowledge of Dutch society. France also tests immigrants abroad ontheir language level, but only as a method to determine whether they shall be obligedto attend an integration course, not to decide if they are allowed to enter France. Inthe near future, three Member States will follow the example of the Netherlands andGermany by adding a language requirement to their admission conditions for appli-cants for family reunification. In November 2010, a language requirement will be in-troduced in the United Kingdom and a language and integration requirement inDenmark. In Austria the introduction of a language test abroad has been announcedfrom October 2, but not yet introducedThe national requirements differ with regard to the organisation and implication,the content and level of the test, the target group and the purpose of the integrationtest. These differences between the Member States will be described below. As thetests involved in family reunification procedures are central, the description startswith the Dutch system.

Basic Civic Integration Examination in the country of residence of the applicant. Theproof of having passed this examination must be handed over on application for ad-mission. Knowledge of the language and Dutch society is tested through an oral ex-amination conducted over the telephone from Dutch Consulates and Embassiesabroad, using voice recognition software which is based in the US. This computerprogramme also decides whether the candidate has passed the examination.Passing the examination is a condition for granting an authorisation for tempo-rary stay, which is for certain nationalities a necessary document for entering theNetherlands. This authorisation is known as ‘MVV’. The migrant must apply for aMVV within one year after passing the examination. After this period, the result ofthe examination becomes invalid and a new test must be taken in order to be admit-ted. If the immigrant fails, he/she will not be granted a MVV, and will thus not beadmitted to the Netherlands. There is no legal remedy with regard to the outcome ofthe examination. The applicant is allowed to do the test as many times as necessary,as long as he/she pays €350 for each examination. The Dutch government does notprovide either courses or learning material. It has however compiled a practice packwhich can be purchased for €70.40 and which includes the film, ‘Coming to theNetherlands’, questions that may be posed during the knowledge of society test, andsome language tests. The costs for an admission procedure for family reunificationare on average €1,440, if the migrant passes the test the first time (€70 for preparationmaterial, €350 for the examination, €830 fee for the visa, €188 for the residence permitgranted after arrival in the Netherlands). These costs do not include the price for aprivate course, which vary (if available) between €450 and €800. As the educationalmaterial is not available in all languages, it is not accessible to all applicants. In onecase, an applicant from Eritrea objected that he first had to learn English in order tolearn Dutch from the educational material. According to the Dutch administrativeCourt however, this complication did not make the requirement disproportionate.After all, the judge argued, the couple could have a family life in Eritrea or Sudan,from where the applicant originated. Despite worrying reports by Amnesty Interna-tional regarding the human rights situation in these countries, the court did not takethe asylum related aspects of this proposed alternative by the government into ac-count.

(ALTE).5If an appropriate language certificate cannot be attained in the country oforigin, the Diplomatic Mission has to ascertain the level of knowledge of the Germanlanguage, for instance by a free ‘hearing’ based on the ‘StartDeutsch 1’test. The ‘StartDeutsch 1’test can be taken in Germany as well as abroad at the Goethe Institute orthe telc GmbH. The Goethe Institute uses its worldwide infrastructure to offercourses and examinations.6Although this test is used as a frame of reference by theembassies, it is not important in which way – independently or on the course – thespouse has achieved the required German language skills (Seveker 2008: 199).7According to the evaluation on the language requirement, the waiting period forattending a course is in most cases not longer than two months. For the average par-ticipant the duration of the course is about four months. Some Goethe Institutes havedeveloped special courses for migrants with a low education or without experiencewith the latin alphabet or learning a foreign language. On the other hand intensivecourses are offered to learn the German language within 5 to 7 weeks (Bundesregie-rung Deutschland 2010: 12-13). The fact that spouses ’all over the world’ are able toattend a German language course constitutes a significant difference from the Dutchact. The German authoritiesfacilitate these courses by safeguarding the offer. Thecosts for the whole procedure are less than half the amount of the Dutch admissionprocedure. Candidates who have attended a language course at the Goethe Institutepay a reduced fee or are exempted from the test fee. The costs for a spouse from Tur-key, for example, in order to fulfil the integration requirement for admission, amountto €660 (€490 course fee, €60 test fee and €60 visa fee, €50 for the residence permitgranted after arrival in Germany).If the price of a Dutch language course is taken into account, the ‘German costs’are less than a third of the ‘Dutch costs’. The test can be taken by the spouse as oftenas required. Just as in the Netherlands, not being able to prove the language skillsimplies no entry to the German Federation. Whereas the proof of passing the Dutchtest is only valid for one year, the language certificate issued by the Goethe Instituteofficially cannot lose its validity. Nevertheless the German embassy may test the lan-guage skills in a visa procedure if the certificate was issued more than a year pre-viously.

‘Start Deutsch 1’ test of the Goethe Institute or the telc GmbH, ‘Grundstufe Deutsch 1’ test of theAustrian Language Diploma (ÖSD) or ‘TestDaF’ of the TestDaF Institute e.V.There are currently 149 Goethe Institutes and ten liaison offices in 91 countries as well as test cen-tres in at least 104 countries. Because of the new requirement the institute’s networks as well asthe licensees’ networks were extended. The test centres in Turkey were also extended from threeto eight (soon nine), additionally, eight test centres were established in Morocco.The ‘way of language acquisition’ is of particular importance for ethnic German applicants. Since1996, they must prove their German language knowledge that must be acquired in a family andbe sufficient for basic communication in German within a hearing organised in the country oforigin.

FranceSince 2007 spouses who apply for family reunification have to meet the requirementsof the integration measures. In order to be granted a visa to enter France and joinhis/her family, the family member has to undergo an evaluation of his/her know-ledge of the language and values of French society in the country of origin. The eval-uation is conducted by the French authorities (OfficeFran§ais pour l'Immigration etl'Intégration),or contracting parties. If the assessment shows a sufficient level ofknowledge, the visa is issued. The applicant will also be exempted from taking partin language sessions on arrival in France within the framework of the welcomingcontract (more on this contract in chapter 3). If the evaluation demonstrates insuffi-cient knowledge, the applicant is invited to attend courses in the country of origin.These sessions, which deal with knowledge of the language and Republican values,do not last more than two months (180 hours). After attendance, a new evaluation iscarried out. If it is successful, the visa is issued and the applicant is exempted fromlanguage lessons after arriving in France. If the knowledge is insufficient, the visa isalso issued but the authorities determine the length of the formation sessions to befollowed in France within the framework of the welcoming contract. Hence, the visais issued dependent upon attendance on the course, but independently from the levelof knowledge the immigrant has demonstrated. Therefore the significant differencebetween the Dutch and German rules and those of France is that according to the lat-ter passing the test is not a condition for the exercise of the right to family reunifica-tion. The test and formation sessions, in the country of origin and after arrival inFrance, are free of charge and financed by theOffice Fran§ais de l'Immigration et del'Intégration.

United KingdomWhereas the previous three countries only introduced a test for family reunification,the United Kingdom started the introduction of the test in 2004 for ministers of reli-gion. Since November 2006 highly skilled migrants seeking to enter the UK as part ofthe Highly Skilled Migrant Programme also have to pass a language test or submitproof of sufficient English education. Since November 2008 all skilled workers haveto meet the language criterion. In summer 2010, the new Conservative-Liberal coali-tion announced that the language requirement will also be applied in applications forfamily reunification from 29 November 2010. If the spouse has to show his/her quali-fication because he/she is not able to submit relevant proof, this will have to be withan oral test provider approved by the UK Border Agency.

family reunification must pass an immigration test in order to be granted a residencepermit on the grounds of family reunification with a spouse/partner in Denmark.8The immigration test is an oral one, consisting of a Danish language test and a societ-al knowledge test. The Danish Act seems to be a copy of the Dutch Integration Act.Besides the similarity in the content and the level of the test, the Danish system hasalso imitated the Dutch view on supporting the preparation for the test. Instead of of-fering courses (abroad or in Denmark), a preparatory package has been developed:the most central part of the package is an educational film, ‘Life in Denmark’.9Theaim of the film is to give the participants a realistic general picture of Denmark andthe daily life in Denmark. Thus, the film will communicate both facts and values witha view to adapting the expectations of the immigrants to the reality of living in Den-mark. Furthermore the package includes a recorded vocabulary list for the languagetest, 100 images from the film with information about Denmark and Danish society,two samples from the language test and test instructions. The packet costs DKK 50plus shipping and administration fees (approximately DKK 150 or €20 ).Although the Danish government was inspired by the Dutch example, the Da-nish application of the requirement differs from the Dutch application in four re-spects. The most important difference is with regard to the location of the test: in-stead of the embassies abroad, the examination will be held in Denmark, after thespouse has received a pre-recognition regarding the fulfilment of the other condi-tions for admission.10Applicants subject to a visa requirement will be granted a visafor three months with a view to taking the test in Denmark. Their stay in Denmarkwill offer them the opportunity to practise the Danish language with their family.11The examination has to be taken within three months after arrival in Denmark; how-ever, it is recommended to take the test within the first two and a half months afterarrival in Denmark in order to leave time for the examiners to assess the test results.A second difference from the Dutch system concerns the preparation possibilities.The Danish educational film is accessible in a Danish and English version on the In-ternet. However, in order to watch the film in any other language (it is recorded in 18other languages) the applicant must buy the preparation package). Thirdly, the valid-ity of the test result is not limited; in the Netherlands, the test has to be retaken if theimmigrant does not comply with the other conditions for admission within 12months after passing the test.A fourth difference from the Dutch system is the method of testing: although the testwill also be arranged by a computer-based test system, the test will be evaluated by8910The Act authorises the Minister of Integration to determine when the immigration test is to comeinto force.‘Livet i Danmark’, see http://www.nyidanmark.dk/da dk/Integration/integration_af_ nyankom-ne/indvandringsproven/Et+liv+i+Danmark+–+undervisningsfilm+til+indvandringsprøven.htm.According to the analysis, around 1,500–2,000 applicants were expected to take the case and mostof them (around 70 %) would already stay in Denmark being issued a tourist visa or another kindof residence permit.On the home page of the Ministry of Integration it is stated that it is the applicant’s own respon-sibility to learn Danish which can be done with the help of the applicant’s spouse or partner, bytaking courses in the country of residence, by buying language courses in the form of books orCDs, or by taking online language courses.

NetherlandsThe required basic level is A1 minus, and the requirements are limited to listeningand speaking skills. The test consists of repeating sentences, answering short ques-tions on basic information, responding to words by saying a word with an oppositemeaning, and retelling a short story. The required knowledge of Dutch society con-sists of ‘elementary practical knowledge’ on the Netherlands, (including geography,history, legislation and political science), housing, education, the labour market, thesystem of health care and civic integration. Furthermore the required knowledgecovers the rights and duties of migrants and citizens in the Netherlands and the ac-cepted norms in everyday life and in society. The knowledge is tested on a level nohigher than A1 minus.The Dutch government has announced that it will raise the level to A1 and ex-tend the language requirements for reading and writing skills at the beginning of2011. With this extension, the Dutch requirements will become more severe than theGerman ones, taking into account that the Dutch examination also covers knowledgeof Dutch society.

four language skills. The written examination contains listening, reading and writing.The oral examination is taken in a group (with a maximum of four): each candidatehas to introduce himself/herself, provide information and ask for information, as wellas make a request and respond to it. Test candidates have to communicate basic in-formation about their name, age, country, address, active working languages, profes-sion and hobbies. They also have to be able to spell their names and to deal fluentlywith numbers. Furthermore, the appropriate use in everyday situations in Germanyby linguistic means is decisive. This means that candidates must be familiar with texttypes such as signs, posters, catalogues, e-mails, postcards, and similar forms. Theymust also possess specific information about the country, culture and everyday life.

NetherlandsThis entry condition applies to those persons aged between 18 and 65 who meet threecriteria: they have applied for admission to the Netherlands with a view to settlingpermanently, they need to have authorisation for temporary stay, and they are ob-liged to participate in a civic integration programme after arrival in the Nether-lands.13In practice this obligation primarily concerns applicants for family formationor family reunification with a citizen of the Netherlands or with a migrant originat-ing from a non-EU country.14Furthermore this act applies to ministers of religioncoming to the Netherlands in order to enter the labour market.There are large groups of migrants who are not affected by the act. These includemigrants who are not required to apply for an authorisation for temporary stay.These are citizens from the Member States of the EU and EEA, Surinam, Australia,Canada, US, Switzerland, New Zealand, Iceland, Japan and North Korea. Also appli-cants with a work permit, the self-employed and highly educated migrants do notfall within the scope of the act. This is also the case for migrants who were granted astatus on the basis of the Long-term Residence Directive (2003/109/EC) in anotherMember State and who fulfilled an integration condition for this purpose. Finally,family members of a migrant with an asylum-related residence permit do not need topass the test, unless the marriage was concluded after the sponsor was granted a res-idence permit (family formation).Migrants who are exempted from the obligation to take the integration test in theNetherlands because of their educational background are also exempted from pass-ing the integration test abroad. These are migrants who have spent eight years ormore in the Netherlands during childhood and migrants with a school diploma orcertificate of education in the Dutch language.Migrants can also be exempted if they are unable to pass the test due to a mentalor physical disability. The legislator refers to the situation where the applicant isblind or deaf or has difficulty hearing, seeing or speaking and is not in possession ofaudio-visual aids. Proof of this disability requires a declaration from a doctor or ex-pert who is appointed by the head of the embassy or consulate. This medical assess-ment takes place at the expense of the applicant. Being functionally illiterate does notconstitute a ground for exemption. Although the government introduced an oral test1314Detailed information on this Act is to be found in chapter 3.Family reunification means that the marriage was already concluded before the applicant wasadmitted to the Netherlands; in other cases (including marriages with Dutch nationals) the defini-tion family formation is used.

GermanyIn Germany, spouses of German citizens or third-country nationals are obliged topass the test abroad. In the interests of close economic relations, the citizens of certaincountries are exempt. These countries are similar to the countries the Dutch govern-ment exempted from the test as well.15Furthermore, spouses of the nationals whomay enter Germany without a visa are exempted from passing the test. Hence, familyreunification for these nationals residing in Germany is easier than for German citi-zens themselves. This is also the case for highly skilled migrants and other migrantswhose residence in Germany is considered to be in the German interest. Further-more, spouses of migrants who have been granted a residence permit for asylum rea-sons do not fall within the scope of the act.Germany has a special regulation in the case of the subsequent migration of chil-dren between the ages of 16 and 18 years who wish to reunite with their parent(s) re-siding in Germany. These children have to prove that it can be expected that theywill integrate into German society. For this purpose it is assessed whether the childpossesses the language ability at CEFR level C1 or if it appears, on the basis of thechild's education and way of life to date, that the child will be able to integrate intothe German way of life.16The integration requirement is not applicable to spouses whose need for integra-tion is discernibly minimal. This is the case for instance if they are in possession of anacademic degree or a comparable qualification or are employed as managing execu-tives, professional sportsmen, journalists or scientists, researchers or teachers. Inpractice, possession of an academic title appears not to be sufficient for exemptionfrom the required German language knowledge. Moreover, employees of an interna-tional company who are based in Germany for no longer than three years, and theirspouses, are also exempt. In practice the requirement is targeted to low educatedspouses from certain non-western third-countries, i.e. Turkey, Kosovo, Russia orThailand. The criteria for exemption on the basis of a physical or mental disability aresimilar to the Dutch criteria. Illiteracy and pregnancy are not grounds for exemption.

FranceThe French legislation does not, unlike the Netherlands and Germany, make a dis-tinction between nationalities of countries from outside the EU. It has laid downthree types of exemptions from the test. The first exemption relates to the age of theapplicant: applicants below 16 and over 65 years old are exempt. The second relatesto educational background, which implies that the requirement is not applicable toforeigners who completed at least three years of secondary studies in a French schoolabroad successfully or who completed at least one year of college (University) inFrance successfully. The third type of exemption involves migrants who have diffi-culties doing the test because of the general situation in their country (war or a natu-ral or technical disaster) or because of personal circumstances, for instance profes-sional obligations or physical or financial difficulties.

DenmarkWith regard to nationality, Denmark follows the French example: apart from citizensof the EU and EEA – and in Denmark also foreign nationals seeking family reunifica-tion with a Turkish citizen living in Denmark who is economically active as an em-ployee, self-employed person or service provider (covered by the 1963 Turkey – EUassociation agreement) - the requirement comprises in principle all foreigners apply-ing for reunification with a spouse or partner (and ministers of religion) (Ersbøll2010: 129-130). Thus, foreigners coming from countries such as the US, Australia, Ja-pan, North Korea are also covered (unlike the case in the Netherlands and Germany).Accompanying spouses of migrants with a residence permit granted for occupationalor educational reasons do not fall within the scope of the act.The test requirement will not apply if the foreigner has previously stayed inDenmark for at least five years and fulfils the Danish language requirement for per-manent residence.Exemption from the test requirement is possible under certain special circums-tances, for instance where the sponsor is a refugee who cannot take up residence inhis/her country of origin due to the risk of persecution or where the sponsor’s per-sonal conditions call for exemption. In all cases where a refusal will constitute a vi-olation of the European Convention on Human Rights (ECHR) dispensation must begranted. In general, the test requirement does not apply to foreigners who cannot ful-fil the requirements due to serious illness or disability, including post-traumaticstress disorder.

NetherlandsThe idea that migrants should start their integration before their departure to theNetherlands, emerged from an evaluation of the integration policy in 2004. The par-liamentary commission (the Blok Commission) which had evaluated the integrationpolicies of the previous 30 years, concluded that the integration of many aliens hadbeen successful (Commissie-BlokBruggen Bouwen:2004). According to the commis-sion, their success was only to a relative extent the result of pursued integration poli-cy. This commission found that 25 per cent of the participants on integration coursesdid not accomplish the desired A2 level. For this disappointing result, the commis-sion pointed to the weak organisation of the integration education: the slow devel-opment of courses, the lack of quality and tailor-made education and the existence oflong waiting lists. Although the evaluation had not revealed a certain unwillingnessof migrants to integrate, politicians immediately blamed the integration policy for itspermissive character. The demand was heard in parliament and government for aradical change in the integration regime by strengthening the responsibility and obli-gations of the migrant. The government announced the introduction of two kinds ofobligations. First, immigrants would be required to pass a basic examination in theircountry of origin as a condition for family reunification. Second, all immigrants whowished to stay in the Netherlands on a permanent basis would have to pass an inte-gration examination within 3.5 years after their arrival. Failing the examinationwould entail financial sanctions and the refusal of a permanent or independent sta-tus. With the decision to start with integration before entry, the government obvious-ly wanted to be able to apply the refusal of entry as a sanction for not fulfilling the in-tegration requirement.The government especially targeted its new policy on non-western spouses of aDutch citizen or third-country national residing in the Netherlands. According to thegovernment, their immigration would cause the largest integration problems. Itstated that ‘the large scale immigration of the last ten years has seriously disruptedthe integration of migrants at group level. We must break out of the process of (fami-ly) migration which time and again causes integration to fall behind’. In particular,the integration process was thought to have been ‘held back by the fact that a largenumber of second generation migrants opts for a marital partner from the country oforigin’. According to the government, ‘an important part of these *family migrants+has characteristics that are adverse to a good integration into Dutch society. Mostprominent among these – also in scale – is the group of marriage migrants from Tur-key and Morocco’ (Bonjour 2010: 306). Almost half of the family migrants would be-long to these communities and would find themselves in a bad socio-economic posi-tion. The government described family migration as a ‘self-repeating phenomenon ofserial migration’ which seemed to be a ‘self-repeating phenomenon of continuousgrowth of ethnic minority groups in a socio-economic deprived position’.The government mentioned four purposes of the introduction of the integrationtest abroad. First, the test would enable family migrants to act more independentlyaf-ter their arrival. Second, it would allow them to make a more deliberate and better21

informed choice on moving to the Netherlands. Third, it would make the migrantand his/her partner residing in the Netherlands more aware of their responsibility forthe integration of the newcomer into Dutch society. The government felt that the‘supply-oriented approach’ was no longer appropriate: emphasising that the own re-sponsibility of the migrant would fit into the new approach to integration that it hadin mind. In this view, supporting the migrant in his/her preparation for the testabroad would send the wrong signal. Furthermore, offering no support would allowthe migrant more freedom of choice in how to prepare for the examination.As a fourth and final purpose of the WIB, the government expected the integra-tion requirement would work as a ‘selection mechanism’: only those with the ‘moti-vation and perseverance’ necessary to integrate successfully in the Netherlandswould be admitted. The government explained it preferred delay or even cancella-tion of family migration to the situation in which integration immediately after arriv-al in the Netherlands would lag behind. It stated that reduction in immigration was‘not a primary goal’, but nevertheless welcomed the ‘side effect’ that WIB was ex-pected to result in a decrease in family migration flows by an estimated 25 per cent(Bonjour 2010: 306).In its Explanatory Memorandum, the government explained how the new re-quirement fitted in with the European developments, inter alia referring to the Fami-ly Reunification Directive and the Long-term Residents Directive. It did not mentionthat the Dutch government itself was the strongest promoter of the insertion of anoptional clause regarding the integration requirements in those directives.17The gov-ernment also considered the integration test abroad to be in line with the ‘Tampereconclusions’, in which the European leaders of governments in 1999 announced thestrengthening of the residence rights of migrants in order to improve their integra-tion (see Tampere conclusions).The bill was highly disputed inside and outside parliament for two main reasons(Van der Winden 2006; Spijkerboer 2007). First, policymakers debated the legality ofhaving a mandatory language test without providing sufficient facilities for immi-grants to learn Dutch since the government was relying on the free market in thecountries of origin to respond to demand for Dutch language instruction. Two offi-cial advisory bodies concluded that the test would violate Article 8 ECHR if a largegroup of family migrants wouldde factobe prevented from living with their spouseor partner in the Netherlands. Second, the linguists Minister for Migration and Inte-gration Verdonk asked to report on the test and the way it was to be administered (ina telephone conversation with a computer) disputed the validity of the languagetest,18as it was based on software developed in the United States for a completely dif-

ferent purpose (Willems 2009: 123-156). The committee Verdonk established to ad-vise her on new integration policy concluded that civic integration could not beproperly tested abroad. Verdonk did not follow this advice and disregarded the opi-nion of the linguists she had consulted. In the end, all major parties, except the GreenLeft MPs, voted for the act, which entered into force in March 2006 (Groenendijk2010: 13).One year after the coming into force of the act, the court judged that the govern-ment was allowed to make the migrant fully responsible for the preparation of theexamination. According to the judge, the legislator had taken these possible obstaclesinto account.Two years after the Integration Act Abroad had entered into force, the NGOHuman Rights Watch urged for the abolition of the civic integration examinationabroad. The organisation deemed the act discriminatory, as it only applied to familymembers from ‘non-western’ countries. As the difference in treatment had no rela-tion to the aim of the measure (better integration in the country of destination), andthe government had failed to justify the difference, Human Rights Watch (HRW2008: 4 and 24-29) considered the distinction to be (direct) discrimination on the basisof ethnic origin and nationality and therefore incompatible with Article 14 ECHR andArticle 26 of the International Covenant on Civil and Political Rights. Although thiscriticism was uttered with regard to the Dutch legislation, it is also applicable to thedistinction the German legislation makes on the basis of nationality. Furthermore,Human Rights Watch argued that the Dutch legislation amounted to indirect racialdiscrimination (and therefore to violation of the UN convention on the Elimination ofAll Forms of Racial Discrimination) because it disproportionally affected residents ofTurkish and Moroccan origin in the Netherlands who wanted to live with theirspouse and children. From the parliamentary debate it appears that the governmentwas especially targeting these two groups. The Social Democratic Minister of Integra-tion replied that the measure was in compliance with European and internationaltreaties. She mentioned three justifications for the different treatment. First, the re-quirement was linked with the existing difference between countries whose citizensdid not need to apply for a MVV and other countries. Second, citizens who were ex-empted because of their nationality were in a cultural, economic and social situationfrom which it could be expected that they had a good understanding of the Dutch so-cial relations, norms and values. Third, the interest in requiring an integration levelfrom them was lower than the Dutch interest in maintaining good foreign and eco-nomic relations with these countries. These interests could be at stake if the govern-ment decided to introduce a MVV and an obligation to integrate before admissionfrom citizens who were currently exempt from this requirement on the basis of theirnationality. The minister added that the Dutch policy served as an example withinthe European Union. A few months before this reply by the government, a courtjudged that the WIB was not discriminatory, because the protection of the economicrelations with these countries justified the ground for exemption. In March 2010, theCommittee on the Elimination of Racial Discrimination (CERD), in its concluding ob-kit costs €65 (about US$80). The fee for the examination is €350 (US$431) and has to be paid eachtime the examination is taken.

servations on the application of the UN convention in the Netherlands, endorsed thepoint of view of Human Rights Watch. The CERD found that the exemption led todiscrimination on the basis of nationality, particularly between ‘Western’ and ‘non-Western’ state nationals, and recommended that the Netherlands review its legisla-tion (CERD 2010).Two critical comments emerged from the Council of Europe. In 2008, the Euro-pean Commission against Racism and Intolerance (ECRI) expressed its concernsabout the reduction in applications and the fees for the examination. It recommendedmonitoring the impact of the test abroad and reviewing the system of exemptions, inorder to comply with the prohibition of discrimination on grounds of nationality(ECRI 2008: nos. 50, 57 and 58). In spring 2009, the Commissioner for Human Rightsof the Council of Europe, Hammarberg, presented his findings on the Dutch policyregarding human rights. In his view, the Family Reunification Directive did not al-low Member States to impose passing an examination as a condition for family reuni-fication. He requested the government to review entry conditions for family migra-tion to ensure that tests, fees and age requirements did not amount to a dispropor-tionate obstacle (Hammarberg 2008: par. 4.2 no 83 and rec. no 19). In its reply, theDutch government agreed with this recommendation and referred to the coming re-sults of the evaluation of the WIB.When it became clear that approximately 90 per cent of the candidates passed thetest, the minister considered the introduction of two strengthening measures: raisingthe limit to pass in order to decrease the number of successful candidates, and in-creasing the minimum test level from A1 minus to A1, which would make the ex-amination more difficult. Both measures were strongly supported by the majority ofthe parliament. The limit to pass has been raised since 15 March 2008, but the minis-ter felt that raising the test level to A1 would only be justified if the governmentwould facilitate the preparation of the test. To this end the minister announced to as-sess the possibility of cooperation with the Goethe Institute, which supported candi-dates worldwide with their preparation for the test on the German language.Statistics showed a significant decrease in the number of applications for familyreunification after the introduction of the integration requirement (the first two years40 percent, later 25 percent. While responding to the evaluation of the integrationrequirement abroad, the government mentioned this drop in the number of applica-tions without judging this consequence of the measures positively or negatively. Itexpressed its concern about the fact that a quarter of the partners still had a pooreducation and that the lasting impact of the integration test appeared to be limited.According to the government, the latter was due to the low level of the examination.It therefore announced that it would raise the examination level to A1 and include awritten examination. The government announced that it would develop ‘specific ma-terial’. With its decision, the government neglected the advice it had requested fromexperts on linguistic integration. The experts recommended not introducing writtentests without proper education, as it would exclude illiterate persons and migrantswho had learnt another alphabet. According to the experts, they would only be ableto pass the test if they could participate in language courses. In order not to changethe way of testing, the Dutch government finally decided to add only a test reading.

GermanyThe language test abroad was introduced as a transposition of one of the non-compulsory conditions of the Family Reunification Directive. The debate showed theclear influence of the politics of other countries: in the legal policy debate about therestrictions on the reunification of spouses, reference was made to the integration re-quirement introduced in the Netherlands. Although the restrictions on the family re-unification of spouses are phrased neutrally in the wording of the law and apply toreunification with German nationals as well as foreign nationals, they are meant toavoid the situation where Turks, in particular, who hold traditional values and whoare living here, bring very young wives uninfluenced by western values from theircountry of origin to Germany. According to the Explanatory Memorandum, the lan-guage requirements are justified by three purposes: the need for promoting or de-manding integration, the aim to provide protection from forced marriages and viola-tions of human rights, and thirdly the need for protection of the social welfare state.19Protection from forced marriages through the introduction of the language test be-fore entry was crucial in public debates and the media. In public discussions in Ger-many, forced marriage is often defined as a human rights question (Ratia & Walter2009). According to the Explanatory Memorandum, in-law families use the lack ofGerman language ability deliberately or indirectly to prevent the victims (mostlywomen) from having an independent social life. The legislator argued that the obli-gation to attend the integration course after entering Germany should not applyequally because of the time delay before the beginning of the course and the processof language learning, while the victim would be subjected to the will of the family-in-law. Furthermore, educated men and women would be more unattractive, accordingto the family concept of affected circles, and would be difficult to ‘control’.After the introduction of forced marriage as a specific offence in 2005, the debateson how to prevent forced marriages focused only on migration law. This culminatedin the two additional new entry requirements – minimum age as well as basic know-ledge of the language prior to entry – for spouses of third-country nationals and19The position of German law regarding Germans with a migration background becomes evenclearer in the justification of economic discrimination, which was introduced with respect to fa-mily reunification with Germans. Concerning the requirements guaranteeing subsistence, a deci-sive factor is whether it is possible to build family unity in the country of origin of a spouse. Thelaw makes a distinction between German nationals: in future, family reunification cannot only bedenied to third-country nationals but also to Germans if the sponsor cannot guarantee a sufficientincome. The former privilege for spouses of a German ceases to apply. Pursuant to the Explana-tory Memorandum, ‘special circumstances’ exist for persons of whom matrimonial cohabitationabroad can reasonably be expected. This especially concerns holders of dual citizenship with re-gard to the country whose nationality they possess in addition to German nationality, or Ger-mans who have lived and worked for a fairly long time in the spouse’s country of origin and whospeak the language of this country.

Germans. Criticism of the constitutional conformity with the provisions and the lackof research on the effectiveness of this restriction and on less general and far-reachingalternatives did not change the dominant view that the protection from forced mar-riages and dependence of brides justified the language requirement. Nevertheless,the bill was only desired and promoted by the Christian-democrats. They succeededin agreement on the bill in a political compromise with their Social-democratic coali-tion partners. The SDP got the consent to regularise a large number of asylum seek-ers who had lived in Germany for years with an insecure status. In return, it had toagree to the Christian-democrats’ plan to introduce a language test abroad (Groe-nendijk 2010).In general, the Courts have already accepted the arguments of the legislator. TheFederal Administrative Court of Germany has confirmed that the regulation is com-patible with the Constitution, the Family Reunification Directive and Article 8 ECHR.The Court decided that the principles of proportionality are applied adequately. TheHigher Administrative Court of Berlin-Brandenburg (OVG) argued similarly regard-ing family reunification with a German citizen. The Federal Constitutional Court ofGermany has not yet ruled on the issue of inferior treatment in cases of family reuni-fication with German citizens compared to EU citizens, regarding the conflict as a re-sult of so-called ‘reverse discrimination’ (Walter 2008). However, the invalid applica-tion of language requirements in the family reunification of EU citizens’ spouses thathad initially been practised was changed as a consequence of the judgment of theEuropean Court of Justice on 25 May 2008 in the ‘Metock’ case (Case C-127/08, cf.8.Lagebericht2010: 475). This confirming jurisprudence is in contrast with the increasinglegal debates in literature on the compatibility of the new language requirements be-fore entry with superior rules of law.

FranceIn France, the debate on introducing obligatory pre-departure integration measuresstarted in 2006, when an MP published a report that made reference to the Dutch in-tegration test and similar measures under discussion in Denmark and Germany. Therelevant bill, introduced in parliament in 2007, makes explicit reference to the provi-sion allowing (but not mandating) integration measures in the EU Family Reunifica-tion Directive. The decision to implement this option to introduce integration meas-ures abroad, was apparently motivated by the same concerns observed in othercountries: the number of third-country nationals admitted for family reunificationwas considered too high in comparison with the numbers of immigrants admitted foremployment or study (Carrera 2009b: 315-316). French President Nicolas Sarkozy(then interior minister) coined two phrases: ‘passerde l’immigration subie à une immi-gration choisie’(‘from passive immigration to selective immigration’ — family migra-tion being in the former category) and ‘uneimmigration choisie, une integration réussie’With the second phrase, the minister claimed a causal relationship between selectiveimmigration and successful integration (Michalowski 2010). Although the integrationtest was only implemented in the Netherlands in 2006, the French minister pointedout the already positive impact of such measures on the integration of immigrants.26

MP Mariani mentioned a second argument for the measure; he pointed to the factthat non-renewal of the residence permit due to failure to comply with the obliga-tions of the welcoming contract was difficult because of Article 8 ECHR. He statedthat such difficulties could be overcome by integration in the country of origin. Inthat situation, if attendance at language courses was not completed, the visa could berefused. Hence, integration abroad was not dedicated to the enhancement of integra-tion of third-country nationals but to better manage migration flows.Although the question relating to integration measures abroad had not been dis-cussed extensively in the parliament, the bill had been significantly changed beforebeing adopted in 2007. Whereas the bill established integration conditions (as the is-suance of the visa was subordinated to attendance at sessions in the country of ori-gin), the adopted law introduced integration measures without the initially proposedrefusal of entry as a sanction for not fulfilling the integration requirement. The origi-nal idea of following the Dutch model and requiring the passing of a language testwas dropped on legal grounds. It was considered probable that the ConstitutionalCouncil, which upon the request of MPs decides on the constitutionality of a bill im-mediately after its adoption, would rule that the requirement violated the FrenchConstitution’s provision for a right to family life. Moreover, it was considered doubt-ful whether this requirement would be compatible with France’s obligations underthe EU Family Reunification Directive, which only allows for integrationmeasuresand not for integrationconditions (Pascouau 2010).

United KingdomThe possibility of introducing a pre-admission language requirement for third-country nationals seeking to migrate to the United Kingdom as the spouse or unmar-ried partner of a British citizen or a person with a permanent residence permit (‘inde-finite leave to remain’) was mentioned in a government policy document publishedin March 2007. A detailed proposal was made in December 2007, setting the requiredlevel at A1. The integration of the spouses/partners into the community and theiremployment prospects were mentioned as the main justifications. In yet another pol-icy document of July 2008, a formal requirement to have basic knowledge of the Eng-lish language was qualified as ‘a medium-term goal’ because of gaps in the availabili-ty of English language courses in countries of origin. In June 2010, the new Conserva-tive-Liberal Democrat government announced its intention to introduce an Englishlanguage test for non-European spouses and partners coming to join a UK citizen ora settled immigrant. They will be required to demonstrate basic English (at the A1level) in an oral test in their home country.

tion of a pre-departure integration test along the lines of the Dutch examination in-troduced a few months earlier. After the act on the introduction of the integrationtest had been adopted in 2007, a working group was set up with a view to making apre-analysis of the implementation of the test. The working group concluded that itwould be very costly to establish a testing system abroad (comprising relatively fewapplicants from many different countries). Accordingly the government and theDPP, in their agreement on the country’s budget for 2008, agreed that the 2007 actwould be amended. The act was amended in spring 2010 and the new examinationswere introduced on 15 November 2010. Cost-benefit considerations played a decisiverole in this remarkable policy change.According to the government, the examination was aimed at increasing thechances of rapid and successful integration of immigrants who would also be betterprepared for integration courses after arrival in Denmark. Furthermore, the immigra-tion test should help in securing that foreigners at the outset took responsibility fortheir own integration and proved their motivation and wish to become part of Da-nish society. The Minister of Integration underlined from the very beginning that thepurpose of the test was not to limit the number of family reunifications – and not tokeep foreigners out of Denmark; the government therefore did not expect a distincti-ve decrease in the number of applications. During the debate in parliament in 2010the minister made it quite clear that the test would be adjusted in such a way that all‘can make it out’.20 She underlined that it was not about ‘integration’, but might beseen as a ‘foretaste’, making it possible for applicants to document their interest inbeing integrated and getting familiar with Danish norms. Based on the test some mi-grants might change their opinion about staying in Denmark – having discovered‘what it is all about’ (learning about Danish sexual morality, etc.). The test was sup-posed to send a signal to newcomers that integration was also about individuals con-tributing actively and engaging in their own integration; likewise, the test was aimedat giving applicants some realistic expectations about their life in Denmark and thepossibilities, requirements, obligations, etc. they would meet.The introduction of the immigration test has received relatively little attention inthe media. During the debate in parliament, it was stressed that the immigration testhad not been invented by the government, it had been, allegedly, successfully im-plemented in the Netherlands – and, according to information received, without ha-ving been blamed for violations of international treaty obligations. With regard toone aspect, the government had probably taken the international criticism on theDutch policy into account. It promised the DPP to examine whether immigrants fromwestern countries could be exempted, as in the Dutch legislation. However in the billimplementing this part of the agreement, no such exemptions were provided.Before the legislative work, a number of NGOs and other organisations and insti-tutions criticised the test for being exclusive, especially taking into consideration thelack of education offers and the high fee which as a whole could make it difficult, ifnot impossible, for poor and/or uneducated applicants to pass. In order to solve someof the alleged problems it was suggested making it possible for migrants to take theknowledge test in their own language. Also opposition parties (including Social-20Oral test and preparatory material that do not imply writing or reading abilities.

AustriaIn January 2010, the Austrian Minister of Interior announced a plan to introducemore restrictive immigration rules, one of them being a language requirement forspouses wanting to join their family members in Austria. The minister explicitly re-ferred to the Dutch case. Some elements of the requirement seem to be similar to theGerman system: the test will include an oral and written test and the level will be A1.It differs however with regard to the intended target group, while spouses of Aus-trian nationals will not fall under the scope of the act (Bundesregierung Österreich2010: 38-39). While the idea is still in the planning stage, it has met with strong criti-cism from Green politicians and civil rights groups. The Austrian Red Cross said thatmandatory language classes in the country of origin would be expensive and unfeas-ible. In some countries, applicants would have to travel long distances to the embas-sies or would face danger doing so (Pop 2009; Groenendijk 2010: 19). On 4 October2010, the Minister of the Interior, Maria Fekter (ÖVP) announced in an interview inthe daily ‘Kurier’, that in the autumn an amendment to the Aliens Law would bepassed demanding knowledge of German at the A1 level as a precondition for immi-gration.

Except for the test applied by the French government, which only serves as a mea-surement instrument, not passing the test means that family members are notgranted reunification with their partners in the Member State. Unless those partnerschoose to move to the country of origin of their family member, they will continue tolive separately. How long this separation will take, depends on the ability of thespouse to pass the examination. If the partner has to take resits, the family reunifica-tion will be delayed. If this inability has a permanent character and he or she is notexempted from the requirement, there are two possibilities to live together: in thecountry of origin of the spouse outside the EU, or in the Member State on a irregularbasis. If the spouse who lives in the Member State is a EU citizen who has exercisedthe right to freedom of movement, they can reunite in his Member State of residenceon the basis of Directive 2004/38/EC.

The study on the effects of the integration and language tests abroad is limited to thethree Member States where the tests are currently applied. In this paragraph a dis-tinction is made between the impact on the number of applications and the impact onthe perception, behaviour and integration of the tests. As the French rapporteur wasnot able to conduct interviews, the second analysis is limited to the Netherlands andGermany.In all three countries, the number of applications dropped significantly after theintroduction of the test abroad. In the Netherlands, the reduction was approximately40 per cent in the first two years after the introduction of the test. Although this dropmay partly be the delayed effect of other restrictive measures introduced in previousyears, several studies indicate that the integration test is the main cause of the de-cline.22Although the number of applications slightly rose again in the years after-wards, the number of applications is still significantly lower than before the intro-duction of the test. In Germany, there was an average 25 per cent reduction in visasfor spouses in the first six months of 2008 compared with the same period in the pre-vious year. According to civil servants of German municipalities, the decrease can beascribed particularly to the introduction of the language test abroad and to the acces-sion of Romania and Bulgaria. The reduction in the number of spouses entering theNetherlands and Germany is the most significant with regard to the countries of ori-gin from where the largest communities residing in the two Member States originate.In the Netherlands the reduction of applications from Turkish and Moroccan spousesis relatively high; in Germany the number of visas issued in Turkey, Kosovo, Russiaand Thailand has declined relatively the most (for instance a decline of 38 per cent ofTurkish citizens).NetherlandsWith regard to the Turkish and Moroccan nationalities living in the Netherlands,statistics show a decrease in the number of marriages with a spouse living abroad.Whereas in 2001 more than half of the Turkish and Moroccan second or third genera-tion migrantsconcluded a marriage abroad, in 2007 only 20 per cent of them chose apartner abroad. They mostly still marry a person from the same ethnic community,but more and more with a person already living in the Netherlands. This develop-ment seems to be connected to the more restrictive admission criteria, especially theincome requirement, the integration test abroad and the age-limit (Dagevos & Gijs-berts 2009: 19). As these groups of immigrants were explicitly targeted by the act, thegovernment’s policy seems to have been successful for this part. This developmenthowever does not explain the whole drop in the number of applications. We do not2122Unless otherwise stated, the figures in this section have been taken from the national INTEC re-ports.Monitor Inburgeringsexamen Buitenland,april 2008, IND, p. 6; IND-rapport,Jaarresultaten2006, DenHaag, maart 2007, p. 3;Jaarrapport Integratie2007, Sociaal Cultureel Planbureau, J. Dagevos en M.Gijsberts, 16 april 2007, p. 316;Korte termijn evaluatie Wet inburgering buitenland, eindrapportage,WODC, Ministerie van Justitie, Januari 2008.

know what the current family situation is of the persons who don’t apply anymore.Did they come to the Netherlands irregularly or do they live separately? Until nowno research has been conducted on their situation, and policymakers do not pay at-tention to this ‘black hole’. Nevertheless the figures show that a number of personsdo not manage to pass the examination, no matter how many times the examinationis taken. Although the pass rate for candidates taking the examination for the firsttime remains on average 89 per cent, the pass rate reduces for migrants who have totake the examination twice or more (72 per cent). The pass rate is lower in the case ofelderly migrants and the lower educated.It is not easy to make a comparison between the data on the Dutch and Germanstatistics because the drop in the Netherlands concerns the number of applications,while the drop in Germany concerns the number of applications granted. In theNetherlands, spouses tend to take the test and apply for admission only if they arequite sure that the visa will be granted. This behaviour might be caused by the highcosts involved in taking the test and paying the fees for the application. The numberof applications granted increased even after the introduction of new and streng-thened requirements. This outcome could be related to the change in population ofthe applicants the statistics show: nearly 75 per cent of the candidates have had anaverage or high education, whereas in 2005 around 53 per cent of the applicants forfamily reunification were illiterate or low educated. Researchers who conducted theofficial evaluation of the integration test abroad called this phenomenon a kind of‘self-selection’ (Odé 2009 and Regioplan 2009). This self-selection seems to affect fam-ily reunification harder than family formation (which means that the marriage is con-cluded after the sponsor was granted a residence permit), as figures show that thenumber of applications for reunification have recovered less than those for forma-tion. This could be explained by the fact that this group of family members probablyincludes relatively more elderly and lower educated migrants (e.g. family membersof victims of war). Hence, although the Integration Act Abroad particularly aimed toaffect young spouses, it seems to have hit other groups harder.

GermanyUnlike the situation in the Netherlands, third-country nationals wanting to (re)unitewith their spouse in Germany do not seem to refrain from application because of thelanguage test. However they are rejected more often than in the Netherlands, mostlydue to their failure of the test. Respondents at the Goethe Institute signaled a certainnumber of spouses who register for the test unprepared. In 2008 the pass rate was onaverage 59 per cent, in 2009 this percentage increased to 64. This increase is presum-ably related to the attendance of future spouses at language courses in Germany.23The language teachers interviewed in this study emphasised an increase in interest inthe language course at level A1 in Germany. Regarding the tests taken abroad, aslight difference is visible between the pass rate of candidates who attended a Ger-23In order to be admitted to Germany, they had mentioned a different purpose for their stay inGermany, in the visa procedure, than language acquisition for family reunification.

man language course at the Goethe Institute (74 per cent) and the pass rate of candi-dates who did not do so (61 per cent). In some specific countries such as Turkey, thedifference is more significant: In Turkey the pass rate in 2008 was 60 per cent (92 percent of the course participants succeeded, 57 per cent of the other candidates suc-ceeded), in 2009 the pass rate increased to 68 per cent (92 per cent of the participantssucceeded, 64 per cent of the other candidates). During the first six months of 2009,one-fifth of the candidates who passed the test had participated in the preparatorycourse at the Goethe Institute. Although the so-called ‘external’ candidates mighthave used other preparation methods (private courses, the Internet, courses in Ger-many), participating in the course specifically preparing for the test seems to haveimproved substantially the chances of passing. According to the evaluation by thegovernment, the significant difference in the pass rate of candidates who attended acourse at the Goethe Institute and the other candidates, reveals a difference in qualityof the courses. The quality of private courses is worse compared to the courses of-fered by the Goethe Institute, or at least unpredictable and they are not preparingadequately on the test (Auswärtiges Amt 2010). In the evaluation, the differences inthe pass rates and the number of resits is related to two factors: the differences in (thequality of) course offerings and the differences in education level of the spouses. Thepercentage of female candidates passing the test proved to be significantly higherthan the percentage of successful male candidates.The evaluation shows that the largest decline in the numbers of visas issued wasin the first six months, when spouses had to prepare to take the test. Although thenumber of visas issued for reasons of family reunification is still lower than beforethe introduction of the test, the government concludes on the basis of its evaluationof the act that the number is now on the level that was to be expected if no languagecondition had been introduced. To substantiate this conclusion, the government re-fers to the fact that the number of applications for family reunification from Turkeyalready declined before the introduction of the test. According to the government astructural decrease in the number of visas issued is in line with the developmentworld wide (Bundesregierung Deutschland 2010: 2-3 and 45-47).24The German rap-porteurs are more cautious in their conclusion: they ascertain an increase in thenumber of family reunifications in 2009, but they notify that the pattern is irregular,as the numbers per quarter differ. This can be influenced by seasonal factors. A com-parison of the same quarters in different years, seems to indicate a recovery to a cer-tain extent of the decline immediately after the introduction of the language condi-tion.25

FranceIn France, the number of visas issued for family reunification in the first six monthsof 2009 was 27 per cent below the level in the same period for 2008, before the newpolicy was introduced. The drop is considered to be the result of the strengthening ofthe rules applicable to family reunification and the introduction of the test. With re-gard to the latter, this factor is probably much less determining compared to theNetherlands and Germany, while passing the test is not a condition for admission toFrance. Failing the test only delays the family reunification for two months at most.Nevertheless, in literature on this topic reference is made to the difficulties migrantsmeet if as a result of the test they have to attend a course. If they live far from theeducation institute, following the course could constitute a financial and organisa-tional burden, despite the fact that the education offered is free (Cournil & Depigny2008).

NetherlandsOn the basis of our interviews, it can be concluded that most migrants are positiveabout learning about Dutch society as preparation for their migration to the Nether-lands. Some respondents emphasised they would also have prepared on Dutch socie-ty without a test. The respondents confirmed that learning the Dutch languageabroad does not seem to substantially contribute to their knowledge of the Dutchlanguage. Preparing for the test itself causes a lot of stress and tension, and takestime and money. Four out of ten respondents interviewed in Turkey were offendedand angry about the requirement. One of them pointed to the fact that Dutch citizensare not obliged to learn Turkish before being admitted to Turkey. Most of the res-pondents were of the opinion that learning the language in the Netherlands wouldbe easier and more appropriate than learning it abroad.All respondents emphasised that preparation for the test would have been im-possible or at least much more difficult without having attended a course. Participat-ing in a course also offers the possibility of getting into contact with other future in-habitants of the Netherlands, and getting more realistic expectations of living there.Female candidates especially seem to benefit from this. These respondents were inthe fortunate position of being able to attend a course. Immigrants lacking this op-portunity also lack these advantages and will face more problems with passing thetest.Preparing for the test constitutes difficulties in some situations, especially be-cause of the lack of support from the Dutch authorities. According to the Moroccanorganisation, migrants living in rural areas have problems travelling to the embassyseveral times (as well as communicating with the embassy from a distance) and find-ing preparation material. Often a course is not available to them. The Dutch RefugeeCouncil pointed to the extremely harsh situation in the (former) war countries. Thefamily members have to travel (twice) through unsafe areas, sometimes to an embas-33

sy settled in another country. They also lack educational material and sometimeseven the Internet or electricity. Also the respondents in Turkey who were able to fol-low lessons, pointed to some negative consequences. According to teachers it wasdifficult for migrants who worked full time to do the course. One had to give up herjob in order to be able to follow the lessons. As the admission procedure had been de-layed, she had already waited for a long time without an income. More respondentsfelt negative about the fact that they had to invest a lot (in time and in money) whilethe outcome of the application procedure was insecure. Also the combination of re-quirements caused stress, as the outcome of the test was only valid for one year. Inone case the partner in the Netherlands had lost his job in the meantime: the migrantwould have to take the test again. Some respondents knew migrants whose relation-ship had broken down as a result of the ongoing problems and frustrations becauseof the test and the whole application procedure.The language teachers in Turkey informed that most of the participants on theircourses there were young and relatively highly educated. One teacher in the Nether-lands noticed that after the introduction of the WIB, he only received highly educatednewcomers. This confirms the observed change in the background of the migrants inthe evaluation of the integration act. According to the teachers and migrants in Tur-key, illiterate persons, elderly migrants and the low educated suffer the most fromthe integration requirement. One respondent and one teacher pleaded for an exemp-tion from a certain age. The Dutch Refugee Council pleaded for an exemption formigrants with psychological problems.Furthermore more than one-third of the respondents interviewed in the officialevaluation of the act, expected that specific groups would have less chance of passingthe examination. They referred to illiterates and low educated immigrants, some ofthem also referred to elderly immigrants. The average age of family migrants, whocome from a country where the integration requirement applies, has droppedslightly since the introduction of the act, while their average level of education hasincreased. According to the researchers this could imply that the act triggers a degreeof self-selection, whereby the elderly and lower educated are being deterred most bythe integration requirement.On the basis of this experience, the researchers of the official evaluation advisedinvesting in specific information for the elderly, illiterate and low educated, in orderto remove their psychological barrier. In their view, the level of the examination issuitable for these groups as well. The advisory committee involved in the evaluationof the act, doubted the compatibility of the integration test abroad with Article 7 (2)of the Family Reunification Directive, as the criteria regarding the proportionality ofthe measure might not be fulfilled. More in general, the advisory commission ad-vised improving facilitation of the preparation for the examination (Regioplan 2009).

marry de facto’. A German language teacher pointed out that the content of the testscauses extra problems for certain groups, as the tasks are related to aspects of dailylife in Germany, which are not familiar to all test candidates abroad.From the immigrants’ point of view, listening is the most difficult discipline inthe test abroad. The spoken language in the recordings for the test and the speech ofofficials in the Diplomatic Mission is regarded as rapid. To some extent, migrants ex-plain the high number of retakes by the fact that the test candidates do not pass thelistening part of the test or they sit the test without being prepared for it. Not prepar-ing for the test seems to be related to a lack of motivation to comply with the re-quirements. According to the respondents at the Goethe Institute, teachers abroadhave difficulties gaining support among the migrants regarding the purpose of therequirement and motivating them to prepare for the test.The German authorities and the Goethe Institute experienced cases of avoidanceas well as attempts at fraud. To improve identity controls, registration for the testtakes place in person and on a different day from the test itself. Sometimes the oraltests and the written tests take place on different days as well. These measures areseen as a burden by those affected because of distances to the examination locationand the financial expense. In some countries (the rapporteurs mention Nigeria),problems regarding authenticity of documents and certificates arise because they arenot registered in the country.The visa procedure is repeatedly criticised in that it lacks transparency and con-stitutes, in addition, a tripwire. The fact that the authorities can repeatedly controlthe language ability during the visa procedure is perceived by migrants as trickeryand an arbitrary measure. In order to combat fraud, migrants were asked at the Em-bassy in Thailand, for example, what colour their blouse was and they had to writedown their name and address, although they had already passed the language test.There were also complaints about spouses who were tested by the Foreigner’s Au-thority a second time after arrival in Germany.Interviewed Germans who wanted their spouse to join them in Germany, per-ceived their position as inferior to EU citizens whose spouses were not subject to thelanguage test: ‘I did not understand it, why am I suddenly in a minority?’ To avoidthe language test before entry, a certain number of Germans take up temporary resi-dence in a neighbouring European country in order to enable the subsequent migra-tion of a spouse as an EU citizen.Migrants and municipal officials in Germany viewed the possibility of learningGerman through courses abroad as positive because a basic command of the lan-guage would help the persons involved to make purchases by themselves, to askquestions independently and to make the newly arrived immigrants more self-confident. The courses also promote social contacts with other spouses. This preventsespecially female spouses getting into an isolated position after arrival in Germany.Not only teachers of integration courses, but also migrant advisory services and mi-grants regard the courses as positive but see the costs and efforts involved in the testas a burden. The respondents of the migrant advisory service did not question theobligation for migrants to learn the German language; however, they have spokenout against the fact that this is bound to the tests in the visa procedure.

The fulfilment of language requirements is associated with strenuous effort, psy-chological burdens, partner stress and family stress: ‘Many people are at breakingpoint over it, which means that I give these couples advice about family reunificationand then transfer them to my colleague in the department of separation and divorce’.From the migrant advisory services’ point of view, this regulation reinforces the im-balance of power between women and men and makes a wife emotionally and finan-cially dependent on her husband. ‘We often had dramatic cases; a girl in Afghanistanhad to go to Kabul, through enemy territory, not only to take the course, but also toapply for a visa. Then she was smuggled across the border. *<+.’The stress that goes along with the language requirement does not seem to corre-spond with the effect on the migrant of having passed the test on language ability.Most of the teachers of integration courses interviewed as part of this study consid-ered that the output of the language test before entry was low and the costs for thoseinvolved were high. Generally, the teachers in Germany do not regard the results ofthe language test at level A1 as significant because of differences in the language abil-ity at the first level of proficiency. They sometimes experience that migrants whopassed the test abroad, are not able to speak a word of German.

Do the Tests Promote Integration?NetherlandsThe purpose of the Dutch test was fourfold: acting independently after arrival, pro-moting a more deliberate and better informed choice to migrate to the Netherlands,emphasising the migrants’ own responsibility and selecting the motivated and per-severing migrants.With regard to the language level, the official evaluation did not reveal large dif-ferences between migrants who took the examination abroad and migrants who didnot do so. The researchers noticed only slightly better listening skills by migrantswho took the examination at their time of arrival in the Netherlands in comparisonwith migrants who did not take an examination abroad. The researchers based thisconclusion on a comparison between these two groups of their level of listening dur-ing the intake soon after admission to the Netherlands. The researchers suggestedthat this difference might also relate to the changing background of migrants.Most of the interviewed migrants in Turkey expected that their preparationwould enable them to act more independently in the Netherlands (see a doctor, goshopping). It is noteworthy that this expectation does not correspond with the expe-rience of the three respondents who had already entered the Netherlands after hav-ing passed the test. The level appears to be too low and the knowledge too soon for-gotten for them to be able to act independently in the Netherlands. Also four out offive teachers in the Netherlands hardly noticed any difference between migrants whodid the test abroad and others. One explanatory factor could be the time that passesbetween the test and the start of a course in the Netherlands. Most of the respondentsargued that learning the language in the Netherlands would be much quicker andmore effective. According to one teacher, knowledge of Dutch society is regarded as36

the most useful education. It helps them to prepare for their stay in the Netherlands.Two teachers pointed to the importance of the contact old participants kept with eachother in the Netherlands, which prevented them from isolation. This advantage is ofcourse only applicable to immigrants who are able to attend a course. As the organi-sation of courses is left to the free market, there are only private courses in the maincountries of origin.The purpose of emphasising the migrant’s own responsibility for his/her integra-tion, seems to have the side effect of making migrants feel frustrated, as they expe-rience a lot of difficulty in meeting the integration criterion without any supportfrom the government. From the interviews we learned that most of them already feltresponsible for preparing for their stay in the Netherlands. The difficulties they hadto overcome were not experienced as proportionate. The absence of any offer ofcourses affected this perception.The purpose of promoting a more deliberate and better informed choice is onlypartly achieved with regard to migrants who attended a course, if the teacher isaware of the importance of promoting realistic expectations of residence in the Neth-erlands. Learning the language from a book does not stimulate a more deliberatechoice. As the organisation and quality of courses is completely left to the free mar-ket, the authorities do not influence this effect.Although respondents often mentioned motivation as a crucial element for meet-ing the integration requirements, the interviews confirm the figures that the testabroad creates selection on age and education. This type of selection differs from thepurpose of the policy makers to select on motivation and perseverance. Given the re-duction in the number of applications and the relatively high educational level of theremaining applicants, one could conclude that the government's planned selectionand reduction in the numbers appears to have been effective.

GermanyThe purpose of the German language test abroad is to promote integration and toprotect against forced marriages. Migrants and migrant advisory services in Germa-ny have repeatedly closed the discrepancy between what is demanded and theknowledge that those involved actually possess after the test abroad. ‘I do not think itis good because people do not speak German after passing the test’. Neverthelessnone of the migrants interviewed emphasised that the language requirements for thesubsequent migration of spouses were easy to fulfil.At the same time, it is not possible to judge whether the test protects those af-fected from forced marriages. It was quite incomprehensible to all the intervieweeshow the language test could prevent forced marriages. Migrants, their spouses andmigrant advisory services have repeatedly felt that the language test does not pre-vent forced marriages. The achievement of this latter purpose is also hard to find out,as figures on the scope of the existence of this phenomen are lacking.According to the evaluation of the government however, teachers abroad had no-ticed in some cases that women deliberately failed the examination in order to avoida forced marriage in Germany. Furthermore, the teachers observed that the atten-37

dance of the courses promotes a greater awareness by young candidates of the dras-tic personal changes involved in their migration to Germany. Learning the elemen-tary aspects of the German language would motivate them to continue language les-sons after arrival in Germany (Bundesregierung 2010: 2).Taking into account the figures on applications and the perception of the respon-dents, one can conclude that the test has a selective effect and is perceived by mi-grants and advisory organsations as an offence against family life. At the same time itis highly doubtful whether the language test promotes integration or prevents forcedmarriages. These administrative procedures and costs on the part of the authorities,as well as the efforts made and expenses incurred by those affected would appear tobe disproportionate to a low success rate, taking into account the low language skillsthe teachers in Germany observe. Integration assistance through language courses isperceived as an advantage, as it prepares spouses for their life in Germany and per-haps motivates them to continue studying the language. The causal connection be-tween the proof of language ability and the claim to family reunification howeverconstitutes the problem.

The Member States can be categorised as the ones that require language skills inreading and writing as well as in listening and speaking (Germany and in futureAustria), and the Member States that limit their requirements to listening and speak-ing skills (Netherlands, Denmark, United Kingdom). The Member States of the firstcategory accompany the written and oral test with an infrastructure of education.This German choice to support migrants in their preparation for the test, expresses itspolicy (introduced in 2005) of promoting and demanding integration. Although thedemanding aspect has gained weight since 2007, the promoting element facilitatesmigrants to fulfil the language requirements.The latter category of Member States, which have only introduced an oral test, donot organise an infrastructure of language courses. This clear division will come to anend at the moment the Dutch government introduces the written test in January2011. No courses in the Dutch language will be offered, despite the conviction of a39

previous minister and the advice of experts that the requirement of written skillswithout the possibility of attending a course, will result in the exclusion of low edu-cated and elderly migrants from family reunification.France also offers courses (and obliges migrants to participate), but only to pre-pare family members for their stay in France. Unlike the other Member States withlanguage requirements, France does not require a certain level of knowledge as acondition for admission. It applies an effort obligation, not a result obligation. Ac-cording to the French rapporteur, the French government wanted to ensure compli-ance with Article 7 (2) of the Family Reunification Directive. The European Commis-sion seems to agree with the French interpretation of the limits of the directive. Ac-cording to the Commission, integration tests can only be admissible if they promoteintegration of the family members, without undermining the purpose of the direc-tive, namely the protection of family life. Whether or not the requirement is propor-tionate and to what extent applicants are supported in their preparation, are deter-mining factors for the admissibility of the test (Commission 2008: 7-8).The systems of the two Member States which will only introduce an oral test, of-fer better preparation possibilities than the Dutch system. Observance of the Britishdiscussion on the decision to apply a language test reveals that the availability ofEnglish language courses was a main consideration. Although the British govern-ment will not organise courses, it can be taken for granted that access to an Englishlanguage course worldwide is easier than access to a Dutch or Danish course. Al-though the Danish government only chose to examine the applicants in Denmark forfinancial reasons, this decision might possibly enable migrants to attend a languagecourse in Denmark. This would probably be much more effective than learning thelanguage abroad.This means that compared with the other Member States that have followed theDutch example, immigrants applying for entry to the Netherlands will face the mostdifficulties in preparing for the examination. This will be even more difficult after theexamination requirements are strengthened (April 2011). At that moment the exami-nation will be the most difficult one in all involved Member States, while the prepa-ration possibilities will be the least.The costs for migrants involved with the language or integration test abroad,show significant differences between the Member States. Taking into account thefees, the Dutch application procedure for family reunification is far more expensivethan those of other Member States. France is placed on the opposite end of the spec-trum, as it does not charge the expenses for the test and courses to the applicant.With regard to the target groups, the Netherlands and Germany make a distinctionbetween third-country nationals. The nationals from ‘Western countries’ (includingJapan and South Korea) do not fall within the scope of their acts. This distinction ismade in order to not harm the economic relations with these countries. The otherMember States however have not followed this example. As a result of the freemovement rights, citizens of the EU and EEA and their familymembers are not af-fected by the language or integration tests. It is worth mentioning that with the ex-ception of Denmark a large number of Turkish citizens are affected by the require-ments, despite the EC-Turkey Association Agreement. The decision of the EU Court40

Political ArgumentsEven though the Member States copied the Dutch ideas, different arguments wereused for the justification of the introduction of the test abroad. Where the Dutch gov-ernment emphasised the improvement of the social position, stressing the migrants’responsibility to integrate and the promotion of a more deliberate choice for theNetherlands, the German government mentioned the promotion of integration andprotection of the social welfare state. As the major justification however it stressedthe importance of the combat of forced marriages. This latter argument by Germanywas copied by the Dutch government as justification for the strengthening of the re-quirements. The Danish government not only copied the Dutch test, but also the ar-guments for its introduction. The motivation differed with regard to two aspects: thegovernment put more emphasis on making migrants familiar with the Danish norms,which could lead to the decision not to migrate to Denmark. Secondly, the Danishgovernment stressed that it was not aiming for a reduction in the number of mi-grants. The British government also did not refer to a reduction in immigrants. TheBritish initiative was first justified by an MP as a measure to combat so-called ‘ghet-toisation’, but the proposal for the test itself formed part of proposals concerningforced marriages. In its announcement on the introduction of the test, the Austriangovernment mentioned the test as a restrictive immigration rule, which implies thepurpose of reducing the number of spouses coming from outside the EU.The Dutch government seems to have changed its arguments in favour of the testabroad over time: at the introduction of the test improving the social position of mi-grants was the central purpose. After the evaluation however, the government subs-tantiated the strengthening of the requirements with the purpose of raising the levelof education of the spouses, with the aim of the protection (of the bride) from thefamily-in-law and the need to combat forced marriages. At the time of introductionof the test abroad, the initial choice of an oral form and a low level were argued withreference to illiterate and low educated migrants: they should also be able to pass thetest. Now that the level is to be raised, this argument seems to have lost its validity.The previous position of the government that the level could only be raised to A1 ifan infrastructure for education was provided has also vanished without any debateor argument. This evolution of arguing in favour of the test abroad, feeds the impres-sion that the test is especially applied as an instrument of immigration.Whereas Germany and Denmark referred to the Netherlands to legitimise the re-quirement, France did so in order to prove that the requirement was effective (al-though the Dutch test had only just been introduced). At its initial proposal, theFrench government was the first Member State to express the explicit aim of reduc-41

ing the number of third-country nationals admitted for family reunification and se-lecting candidates on their chances to integrate successfully. Although these purpos-es were to a certain extent similar to those of the Netherlands, the government de-cided to abandon the introduction of passing the test as a condition for admission be-cause of incompatibility with the French Constitution and the Family ReunificationDirective.The Dutch legislation did not convince the French civil servants and politiciansof the lawfulness of the measure, but it seems to have convinced the German andDanish ones (after all there is a rule of law in the Netherlands). The wide internation-al criticism of the Dutch test abroad had not been included in the debates in theseMember States, except for the criticism of the distinction between nationalities. Whe-reas this criticism did not affect the Dutch rules, it may have resulted in the absenceof distinction between nationalities in Denmark. With regard to other elements of thecriticism, the introduction of the test abroad in Germany, Denmark, and to a certainextent the United Kingdom and Austria, was legitimised by the previous introduc-tion in the Netherlands without having taken the critical international reports on theDutch practice into account.Although the Member States which introduced a language or an integration testabroad, seemed to have been inspired by the Netherlands, they all made differentchoices on its implementation. Besides the already mentioned choices regarding theassistance in learning the language, the difference in target group is noteworthy. Thefact that France and Denmark did not follow the Dutch and German example to dis-tuingish certain nationalities, indicates that they did not want to risk a violation ofthe prohibition on discrimination on the basis of nationality. At the time the act wasdiscussed in Denmark, much international criticism was uttered on the distinctionthe Netherlands made between nationalities in applying the integration act abroad.By exempting Turkish citizens from the integration requirement, Denmark hasshown that it has taken the jurisprudence (Commission against the Netherlands) ofthe Court of Justice on the Association Treaty between Turkey and the European Un-ion into account. Hence, international criticism and jurisprudence directed at theDutch policy has not led to changes in the Dutch integration requirement, but it hasactually influenced the Danish integration requirement. The distinction the UnitedKingdom makes emerges from the language that is used in the countries of origin.This can be considered as an exemption because of sufficient knowledge of the lan-guage.

adjust to the new requirement. Furthermore, preparation for the test takes time,which partly explains the sharp drop immediately after the introduction of the test.In the Netherlands more ethnic Turks and Moroccans marry a person already resid-ing in the Netherlands, partly as a result of the cumulation of more restrictive admis-sion conditions. The figures of the Netherlands show that the remaining spousesfrom outside the EU are younger and more highly educated than the population ofspouses before the introduction of the test. The statistics on applicants for familyreunification in the German report do not include figures on educational backgroundand age, but the German respondents also estimated that the language requirementselects on education and age. The selecting effect of the test is probably less signifi-cant in Germany than in the Netherlands, where applicants for the German test aresupported in their preparation by a course.The conclusion of respondents in the Netherlands and Germany are quite similaron a number of aspects. First, they share the opinion that the language requirementonly results in a slight improvement of the language skills of the spouses. Taking intoaccount the stress, time and money that are involved in passing the test, many res-pondents think there is no balance between the efforts and the result. Probably thisimbalance is partly caused by the fact that learning the language of a country whereyou have never been is much more difficult than learning the language when youpractice it in daily life. All respondents in both Member States acknowledged thatlearning the language after arrival is much more effective. Another reason is thatsome of the applicants are not used to learning (a language). The differences in passrates in Germany also show that passing the test partly depends on the quality of thecourses. Introducing a uniform quality mark or certification would perhaps improvethe overall quality of the courses. The preparation for the Dutch test is even harderbecause of the lack of courses or the dependence on private courses without a qualitymark. This will constitute an even larger problem after the introduction of the writ-ten test. Organising an infrastructure of education abroad would diminish the risk ofexclusion of certain groups. Furthermmore it would encourage social contacts afterarrival and result in more realistic expectations of life in the Netherlands. Hence, twopurposes expressed by the Dutch government of the introduction of the test can bebetter achieved by personal education abroad.Another similarity is the problems migrants meet when they live in (post) warcountries like Afghanistan. The absence of an embassy or education institute forcesthem to travel through unsafe areas several times. Also the absence of electricity orother elementary infrastructures constitute obstacles to the preparation for the test.These were probably the reason for the French government to exempt migrants liv-ing in these areas from taking a test and attending a course.In general, migrants we interviewed were positive about the possibility of pre-paring for their stay in the Netherlands or Germany. Women seemed to be more mo-tivated to learn and more positive about the requirement than the men. The oneswho were able to attend a course (of good quality) gained social contacts, more self-confidence and a more realistic expectation about their future life. According to mi-grants preparing for the Dutch test, preparing on society was more useful than learn-ing the language abroad. There was a broad consensus for the view that migrantsshould be required to learn the language of their new home country. A large number43

of them thought, however, that this requirement should not be imposed before de-parture. According to most respondents however, the largest problem was the linkbetween acquiring a certain level of knowledge and admission. A substantial propor-tion of respondents pointed out that the absence of individual considerations whileapplying this general requirement, sometimes resulted in distressing situations.In cases where it is difficult for the migrants to take responsibility, one may ques-tion whether the effort which is demanded is proportionate in relation to the right tofamily life and the purposes of the test. Regarding the question whether the purposeof the test is achieved, the research so far makes clear that the effects on languageskills are marginal and the effects on integration are unknown for two reasons. First,the tests have been introduced only recently, and getting integrated takes time.Second, research on integration shows that the level of integration depends on sever-al factors. An effect which is already evident is the reduction in the number of appli-cations, especially from lower educated and elderly migrants. The tests serve as a se-lection based only on education and age, instead of the intended selection based onmotivation or the combat of forced marriages. Both Dutch and German governmentsdo not seem to assess accurately whether the tests are effective in relation to theirgoals, or take other effects for granted. The fact that the Dutch government decidedto introduce a written test because (inter alia) 25 per cent of the applicants is (still)low educated implies that the actual effect of the test (higher percentage of highlyeducated spouses) has now become its purpose. This shift in the argumentation forthe application of the test abroad makes it more difficult to assess the effectiveness ofthe measure.In the Netherlands and Germany however, the highest national courts have ac-cepted the arguments for the tests and judged that the requirement is in compliancewith the Constitution (Germany), Article 8 ECHR (both Member States) and the Fam-ily Reunification Directive (both Member States). In both countries this judgment isalso made with regard to illiterates. In the Netherlands this was also the case regard-ing a migrant who failed the test four times, as well as a migrant who was forced tofirst learn English in order to read the educational material, as this was not availablein his mother tongue. Until now, no preliminary reference has been made to theCourt of Justice.

Currently, immigrants are required to pass an examination in seven of the nine coun-tries in this study. Austria, Denmark, France, Germany, Latvia, the Netherlands andthe United Kingdom have made access to permanent residence conditional upon thepassing of a language (and knowledge of society) examination for particular catego-ries of immigrants. In Belgium, the Flemish Region requires newly arrived immi-grants to attend an integration programme, but this is not a condition for access topermanent residence. In Hungary, immigrants are neither required to pass an ex-amination nor to attend a course as a condition for access to permanent residence.The requirements differ also with regard to, e.g., content and level, target groups andexemptions, costs for the immigrant, and sanctions for non-compliance. The re-quirements are summarised in table 3.1 and briefly discussed below.Table 3.1: Integration tests after admission to the countryCountryAre immigrantsrequired to passan examination?Are they (also)required to at-tend an integra-tion programmeor course?Formally, noYes, but onlyin the FlemishRegionYesIs knowledge ofthe languagetested (requiredlevel*)?Yes (A2)n.a.Is knowledge ofsociety tested?Year of entry in-to force

Yes (B1 or, ifYes2007they haveopted for at-tending acourse,progress of atleast one level)* The language proficiency levels referred to are those set out in the Council of Europe’s ‘CommonEuropean Framework of Reference: Learning, Teaching, Assessment’. A person with level A1 profi-ciency is defined as: ‘Can understand and use familiar everyday expressions and very basic phrasesaimed at the satisfaction of needs of a concrete type. Can introduce him/herself and others and can askand answer questions about personal details such as where he/she lives, people he/she knows andthings he/she has. Can interact in a simple way provided the other person talks slowly and clearly andis prepared to help.’ A person with level A2 is defined as: ‘Can understand sentences and frequentlyused expressions related to areas of most immediate relevance (e.g. very basic personal and family in-formation, shopping, local geography, employment). Can communicate in simple and routine tasksrequiring a simple and direct exchange of information on familiar and routine matters. Can describe insimple terms aspects of his/her background, immediate environment and matters in areas of immedi-ate need.’ A person with level B1 is defined as: ‘Can understand the main points of clear standard in-put on familiar matters regularly encountered in word, school, leisure, etc. Can deal with most situa-tions likely to arise whilst travelling in an area where the language is spoken. Can produce simpleconnected text on topics which are familiar or of personal interest. Can describe experiences andevents, dreams, hopes and ambitions and briefly give reasons and explanations for opinions andplans.’

AustriaSince 1 January 2003, Austria has required that non-EU/EEA immigrants fulfil an ‘in-tegration contract’ during their first years of residence in the country. Compliancewith this requirement is a condition for the renewal of a residence permit or the is-suance of a permanent residence permit. One of the ways to fulfil the integration con-tract is to successfully complete a German language and integration course at a certi-fied institute. In 2005, the required level of language proficiency was raised from A1to A2. At the same time, the time period within which the integration contract had tobe fulfilled was extended from four to five years. Since 2005, the language and inte-gration courses have consisted of two modules. The first module (75 hours) aims atimparting reading and writing skills to illiterate immigrants; the second module (300hours) aims at imparting a basic knowledge of German (level A2). About a quarter ofthose who rely on a language and integration course first attend the literacy module.The course is completed by a written and oral language examination, held by theteachers and graded by the teachers as ‘pass’ or ‘fail’. The integration contract can al-so be fulfilled without attending a German language and integration course, by pass-ing a language examination at level A2 at a certified language school. The govern-ment intended to increase the required level of language proficiency from level A2 toB1, but it seems to have abandoned this idea. The Minister of the Interior recentlyannounced, however, that the time period within which the integration contract hadto be fulfilled would be reduced from five to two years.

BelgiumSince 2003, newly arrived immigrants in Flanders are entitled and some are obligedto follow an integration programme. The obligation is limited to attendance at thecourses and there is no official examination to assess the end result. Immigrants whohave attended 80 per cent of the courses receive an attestation. Immigrants who be-long to the target group are required to sign and fulfil a ‘contract of civic integration’in which the content of their integration programme is specified. The first, obligatory,part of the programme consists of a Dutch language course, an introduction to Fle-mish and Belgian society, and career guidance. The level of language proficiency tar-geted in the language course is A1. The second part of the integration programme isonly available upon completion of the first part and seeks to help immigrants to fullyparticipate in Belgian society. The programme normally does not last more than ayear.

so passed the language examination at the end of the programme. In 2007, the re-quired level of language proficiency was set at B1. In addition, it was required thatthe applicants had had employment for at least 2.5 years within the last seven years.The language and employment requirement together were labelled ‘the integrationexamination’. The current requirements, which have been in force since June 2010,include a strengthened employment requirement (2.5 years within the last threeyears) and furthermore they include an ‘active citizenship’ requirement (which canbe fulfilled by passing a special ‘active citizenship’ test or through active participa-tion in an organisation for at least twelve months) and a requirement to fulfil ‘sup-plementary conditions relevant to integration’ (which can be fulfilled by having hadfull-time employment for at least four years within the last 4.5 years, having com-pleted certain forms of education, or having passed a Danish language test at levelB2). The ‘active citizenship’ test, which will be introduced in mid-2011, will resemblethe naturalisation test. However, the level will be somewhat lower and it will consistof only 15 questions, ten of which must be answered correctly. Newcomers are en-titled to three years of Danish language tuition. The municipalities have to offer eachnewcomer an integration programme. Its content should be based on an assessmentof the newcomer’s skills, background and needs and must be laid down in an indi-vidual contract. The programme includes a language course. The language coursesare offered at three levels: A2, B1 and B2 (targeting foreigners according to their pre-vious schooling). Each course lasts three years and each course is completed by anexamination. By the 2010 amendment of the Aliens Act it was decided that the inte-gration programme also has to include an ‘active citizenship course’.

FranceSince 1 January 2007, non-EU immigrants who intend to stay in France for more thanone year have been required to sign and fulfil a ‘welcoming contract’. Depending ontheir level of language proficiency, they may be required to attend a language course.If the level of language proficiency has not been evaluated in the country of origin(see chapter 2), it will be determined after the immigrant’s arrival in France, duringthe course of the interview where the welcoming contract is signed. The languagetest consists of a multiple choice exam and an oral and written proficiency test (Car-rera 2009a: 338). If the immigrant does not have the required level of language profi-ciency, he will be required to attend a French language course as part of the welcom-ing contract. The course lasts a maximum of 400 hours and aims at imparting a basiclevel of French (A1.1, which is lower than A1). Students who pass the examination atthe end of the course receive theDiplôme Initial de langue fran§aise.The welcomingcontract also comprises civic formation (6 hours) and an information session aboutdaily life in France (1-6 hours). All training and information sessions are organisedby the Office Fran§ais de l'Immigration et de l'Intégration (OFII). The welcoming con-tract has to be fulfilled within a year. It is considered fulfilled when the immigranthas attended the training and information sessions and acquired theDiplôme Initial delangue fran§aise.Only about one in four or five newcomers are required to attend alanguage course; a large majority of the ‘newcomers’ in France have sufficient French48

GermanyAlready in 1990, Germany started to require a basic knowledge of the German lan-guage from applicants for a permanent residence permit. The Zuwanderungsgesetz,which entered into force in January 2005, provides that non-EU immigrants in Ger-many can be required to successfully complete an integration course as a conditionfor the issuance of a permanent residence permit. The current integration course con-sists of a basic and an advanced language course (600 or 900 teaching hours) and anorientation course (45 teaching hours). The language course aims at imparting an in-termediate level of German (B1). The orientation course aims at imparting a basicknowledge of the legal system, culture and history. Both the language and the orien-tation course are completed by an examination. The current language examinationhas been in use since July 2009. It is a tiered examination, testing not only level B1 butalso level A2. The test is passed when B1 has been achieved, that is, when three of thefour parts (speaking, listening, reading and writing) have been passed with B1. Can-didates who achieve level A2 receive a certificate of their language skills, and obtainthe opportunity to take the advanced language course again (up to 300 hours) toreach level B1. The examination currently in use at the end of the orientation coursewas introduced in January 2009. Candidates have to answer at least 13 out of a totalof 25 questions correctly to pass this test. The 25 questions are derived from a cata-logue of 250 questions based on the three modules of the curriculum of the orienta-tion course.

NetherlandsIn 1998, the Netherlands introduced obligatory integration courses for newly arrivednon-EU/EEA immigrants. In January 2007, a new law entered into force which re-quires not only ‘newcomers’ but also immigrants who settled in the country before2007 to pass an integration examination. In January 2010, passing the examinationbecame a requirement for permanent or independent residence status. Newcomershave to pass the integration examination within 3.5 years from the issuance of theirfirst residence permit. The examination consists of a practical examination and a cen-tral examination. The practical examination tests the language skills of the candidatesin daily life situations. Candidates can pass this examination by submitting a portfo-lio (containing evidence of 20 situations in which the candidate has had to speakDutch, e.g., registering a birth), taking part in an assessment (an oral examinationduring which the candidate has to take part in four role-plays), or by a combinationof both. The central examination consists of three parts. The ‘spoken Dutch test’ is atelephone test in which the candidate has to repeat sentences, answer questions andgive brief accounts of stories. The ‘digital practical test’ and ‘knowledge of Dutch so-ciety test’ are taken using a computer. The digital practical test consists of questionsabout daily life situations. The knowledge of Dutch society test consists of about 43questions about work and income; manners, norms and values; housing; health andhealth care; history and geography; authorities; polity and the constitutional state;and education and upbringing. The level of the examination is A2 (or, for settledimmigrants, A1). Immigrants who are already proficient in Dutch can prove this bypassing a ‘short exemption test’. The level of this test is higher than that of the inte-gration examination, B1 instead of A2. Another route has been created for highlyeducated immigrants: they can take the ‘NT2’ (Dutch as a second language) state ex-amination, which has a higher level than the integration examination and givesaccess to higher vocational education and university.

Those who are already proficient in English will take the ‘Life in the UK’ test, whichis pitched at level B1. It is a computerised test consisting of 24 multiple-choice ques-tions about national institutions, society, employment matters and everyday matterssuch as housing, money, health and education. It is based on an official handbook.The test outcome is communicated automatically to the UK Border Agency. Thosewhose level of English is lower than level B1 will need to attend combined ESOL andcitizenship classes. There is no set format for these courses. The courses are generallyprovided by further education colleges (run by the government of each of the fourcomponent parts of the United Kingdom) or by adult education services (providedby local authorities). Applicants using this route must demonstrate progress of atleast one level to obtain the required ‘ESOL Skills for Life’ qualification. Relativelyfew immigrants use this route. Since the introduction of the language and ‘Life in theUK’ requirements, 81 per cent of applicants for indefinite leave have met the re-quirements by passing the test; 12 per cent have relied upon an ESOL qualification;and 8 per cent have been exempted.

In five of the seven countries where immigrants are required to pass a language orintegration examination, there are publicly regulated and/or funded courses thatprepare them for this examination. In Denmark, France and Germany, immigrantswho belong to the target groups are (or can be) obliged to attend these courses. InAustria, formally there exists no obligation to attend these courses; immigrants whoare required to prove their language skills can also do this by passing a language ex-amination at a certified institute. In the Netherlands, the government sought to pri-vatise the integration courses. The idea was to make the immigrants themselves re-sponsible (also financially) for acquiring the required knowledge and skills; the gov-ernment would only define the requirements. However, it soon became clear that thisdid not work, and the municipalities were again made responsible. They can requireimmigrants who belong to the target groups to attend a course. The centre-right gov-ernment that came into power in October 2010 intends to change the system again.Two other countries with examinations do not provide public courses to preparefor them. In Latvia, no official courses are provided to prepare for the Latvian lan-guage test which applicants for permanent residence must pass. Immigrants have toprepare themselves by studying a book or by taking lessons with a certified teacher.In the United Kingdom, there are no official courses to prepare for the ‘Life in the UKtest’. Immigrants have to prepare themselves by studying the official ‘Life in the UK’handbook. However, immigrants who do not have sufficient language skills are of-fered another route. They can attend combined language and citizenship classes at anaccredited institute and, instead of taking the ‘Life in the UK’ test, submit a formalletter from the institute in question, setting out their initial and final levels reached.The Flemish Region of Belgium does not require immigrants to pass an examination,but it does require them to attend official integration courses.

As the following table shows, there are large differences in the costs of the examina-tions and courses for the immigrants. Whereas the language courses in Denmark andFlanders and all training and information sessions in France are free of charge forimmigrants who are obliged to attend them (and in Denmark also for other residentimmigrants), immigrants in other countries have to pay at least part of the coursecosts themselves. Immigrants in Germany pay €650 and immigrants in Austria pay€750 to €2,500 for their language and orientation courses. Both in Germany and Aus-tria, immigrants who pass the examination within a given time period can claim partof their contributions back. In the Netherlands, the municipalities may demand thatimmigrants contribute €270 to the costs of their integration courses (including ex-amination fees). The costs for immigrants who prepare for the integration examina-tion on their own are much higher. The examination fees also vary widely. Candi-dates in Denmark normally do not pay a fee.Table 3.2: Costs of examinations and coursesAustriaImmigrants have to pay fees for the courses they attend to fulfil their integra-tion contract. The literacy module normally costs €350, the German languagemodule €750 to €2,500. If the first module is successfully completed within oneyear, the costs are reimbursed. If the second module is successfully competedwithin two years (or three years for those who first completed the literacymodule), half of the costs (up to a maximum of €750) are reimbursed. Personswho only take the language exam pay a fee of €50 to €100.The language courses are free of charge for all participants (compulsory as wellas voluntary participants).Municipalities are required to offer newly arrived immigrants free languagetuition for up to three years. Candidates who want to take a language examwithout having attended a course may be required to pay a fee of about €130.A cost-based fee will be charged for the ‘active citizenship test’.All training and information sessions are financed by the Office Fran§ais del'Immigration et de l'Intégration (OFII).Immigrants who attend a language and integration course pay €1 per teachinghour; the total costs are normally €645. Recipients of welfare or unemploymentbenefits can apply for an exemption. Immigrants who pass the integration ex-am within two years can claim half of their contribution back. The test fee (forthose who take the exam without having attended a course) differs per federalstate; in most states, it is between €95 and €125.n.a.The fee for the language test is 10 LVL (about €14).

CHAPTER3: INTEGRATIONTESTS IN THECOUNTRYNetherlandsImmigrants who are offered a course by their municipality can be required tocontribute €270 to the costs of their course and exams. For others, the costs aremuch higher. The fees for the three parts of the central exam are €126 in total.The fee for the practical exam is €104 (portfolio) or €250 to €1,200 (assessment).Those who choose to follow a course will also have to pay a course fee. Thecosts are partly reimbursed if the integration exam is passed within 3.5 years. Itis also possible to get a loan.The fee for the short exemption test is €81 and the fee for the NT2 state exam€90.The fee for the ‘Life in the UK’ test is £34. The official ‘Life in the UK’ hand-book costs £9.99 in hard copy. The costs of the ESOL courses (for immigrantswhose level of English is below B1) vary depending on the provider and thekind of course. There are exemptions from fees for, e.g., persons in receipt ofsocial benefits and persons in humanitarian categories, and reduced fees forpersons protected by EU free movement law and the partners of settled per-sons who have been resident in the United Kingdom for one year.

Seven countries (plus the Flemish Region in Belgium) require non-EU/EEA immi-grants to pass an integration test and/or to complete an integration programme. Thetarget group has been defined differently by each country. However, it always in-cludes newly arrived immigrants who have been admitted for family reunification,and in most countries, a large majority of the target group falls into this category.EU/EEA citizens and their family members are always excluded from the targetgroup. The family members of own nationals, on the other hand, are always in-cluded. In the United Kingdom, however, the parent, grandparent or other depen-dent relative of a British citizen or settled person are exempt from the ‘Life in the UK’test.Immigrants who have been granted asylum do not always belong to the targetgroup; they are not required to pass an integration test or complete an integrationprogramme in Austria, Latvia and the United Kingdom. (In the Flemish Region inBelgium, on the other hand, asylum seekers who have stayed in the country for morethan four months have to attend a course on Flemish and Belgian society.)Many labour migrants in all countries with integration requirements will nothave to fulfil these requirements because their stay is (assumed to be) of a temporarynature. The Austrian legislation contains an explicit exception for highly skilled la-bour migrants (‘key personnel’): non-EU/EEA immigrants who intend to stay in Aus-tria for more than 24 months have to sign an integration contract, but ‘key personnel’(and their family members) are regarded as already having fulfilled the integrationcontract. France has a similar exception for holders of a ‘skills and talents’ visa; theyare not required to fulfil the welcoming contract. The Dutch legislation contains anexception in the opposite direction: ministers of religion are the only category of (as-sumedly) temporary migrants who are required to complete an integration course inthe Netherlands. A similar exception exists in the Flemish Region of Belgium, whereall labour migrants with the exception of ministers of religion are exempt from theobligation to attend an integration programme. In Denmark, the Integration Act was53

amended in 2010 to include labour migrants (including EU/EEA citizens) on a volun-tary basis. The reason for this amendment was that the number of migrants admittedfor employment had more than tripled since 2001. In the United Kingdom, Turkishbusinesspersons recognised under the association agreement with Turkey are exemptfrom the ‘Life in the UK’ test. The other countries do not have exceptions for Turkishcitizens. However, two Dutch courts recently ruled that imposing integration re-quirements on Turkish workers is contrary to the non-discrimination provisions andthe standstill clauses of the association agreement with Turkey. The Dutch govern-ment has lodged appeals against these decisions.In most countries, immigrants who can prove their integration or knowledge ofthe language with particular diplomas or certificates are not required to take the ex-amination. In most countries, young immigrants who are still in education are notrequired to take the examination either. Most countries also have age limits. Severalcountries have limited the target group to persons aged between 18 and 65; othercountries have exemptions for older persons or pensioners and/or for younger per-sons.In addition, most countries have exemptions for disabilities or long-term illnessesthat severely restrict the ability to speak or learn the language or to prepare for theintegration test. The formulation of this exemption in the Danish legislation (handi-capped persons may be exempted from fulfilling requirements, which they are notable to fulfil, provided that it is required by ‘Denmark’s international obligations, in-cluding the UN Convention on the Rights of Persons with Disabilities’) has arousedmuch criticism, because it leaves wide margins of discretion to the immigration au-thorities. In Germany, immigrants who have to look after a handicapped familymember can also apply for an exemption.The United Kingdom also has exemptions for, e.g., victims of domestic violenceand foreign nationals discharged from the armed forces. In Denmark, foreigners‘with strong ties to Denmark’ are exempted from the new ‘active citizenship’ and‘supplementary conditions’ requirements; this exemption applies to foreigners be-longing to the Danish minority in South Schleswig, former Danish citizens, foreign-ers with Danish parents, and Argentinian citizens with Danish parents or grandpa-rents.In Germany, the Netherlands and the Flemish Region in Belgium, not only newlyarrived immigrants but also immigrants who have lived in the country for a longtime (and who may already have a permanent residence permit) can be obliged topass an examination or to attend a course if their language skills are considered to beinsufficient. In Germany, this applies to foreigners who receive unemployment bene-fits and to foreigners who have ‘special integration needs’. The latter category in-cludes parents of minor children living in Germany who are dependent on social as-sistance. In the Netherlands, it applies to foreigners who do not have a diploma prov-ing that they have sufficient knowledge of the Dutch language; they can be obliged toattend a course and to pass an examination (at level A1) even if they have a job.Moreover, having attended an integration course under the previous legislation(which required newcomers to attend a course targeted at level A2, but which didnot require them to pass the examination at the end of the course) is no ground for

SanctionsIn seven countries (plus the Flemish Region in Belgium), non-EU/EEA immigrantsare required to pass an integration test and/or to complete an integration pro-gramme. The consequences for immigrants who fail to comply differ from country tocountry and may range from not being eligible for permanent or independent resi-dence status to being threatened with expulsion. Failure to comply may also lead tofinancial penalties (administrative fines, withdrawal or cutting of social benefits) inhalf the countries studied.Table 3.3: Consequences of failure to pass test or to attend courseCountryConsequences for Consequences forAdministrativeentitlement torenewal of tem-finepermanent resi-porary residencedence permitpermitAustriaYES(YES)YESBelgium (Flanders)NONOYESDenmarkYESNONOFrance(YES)(YES)NOGermanyYES(YES)YESLatviaYESNONONetherlandsYESNOYESUnited KingdomYESNONO(YES) between brackets means that the consequences are not straightforward.Consequences forsocial benefits

In Latvia and the United Kingdom, immigrants who are not able to prove that theyhave sufficient knowledge of the language (and of ‘Life in the UK’) are not eligiblefor permanent or long-term residence status and the rights attached to it. They willhave to apply for renewal of their temporary residence permit. In these countries,there are no other penalties.Denmark and the Netherlands have also made access to permanent residencestatus (and the rights attached to it) conditional on the passing of a language (andknowledge of society) test, but, in addition, these countries have financial penaltiesfor immigrants who fail to comply with the obligation to complete an integrationprogramme. In both countries, these immigrants’ social benefits may be cut. In theNetherlands, the failure to complete an integration programme may also be punishedby an administrative fine. An evaluation study in the latter country found that half ofthe municipalities surveyed were reluctant to impose such a fine, because they didnot think it would help or because they considered it too heavy a sanction for immi-grants with a low income. Other municipalities, however, considered the maximumfine too low (Significant 2010). The centre-right government that came into power in55

October 2010 intends to make it possible to withdraw the temporary residence per-mits of immigrants who do not pass the Dutch language and integration test.In Germany, there is a certain reluctance to emphasise the obligatory nature ofthe integration programmes and tests (cf. Joppke 2007). The ‘right’ to participate isstressed. Many newcomers are ‘entitled’ as well as ‘obliged’ to enrol on an integra-tion course. Those who pass the test at the end of the course will be granted a per-manent residence permit (provided that they fulfil all other requirements for such apermit). However, not having passed the test will not automatically lead to the refus-al of a permanent resident permit. It will depend on whether the immigrant in ques-tion has attended the course ‘properly’. On the other hand, gross and repeated failureto comply with the obligation to attend a course may also, although again not auto-matically, have consequences for the renewal of the temporary residence permit. Asthe immigrant’s duration of stay in Germany, his/her ties with Germany, and theconsequences for his/her family members must be taken into account, this penaltywill rarely be imposed. However, failure to comply may also be punished by an ad-ministrative fine or the reduction or withdrawal of the immigrant’s benefit.Austria has made access to permanent residence status conditional on the suc-cessful completion of an integration contract (or the passing of a language test). Theintegration contract has to be completed within the first five years of residence, butthis period may be extended, on request, for a further two years (and the extensionmay be renewed). An immigrant’s failure to fulfil the integration contract may alsohave consequences for the renewal of his/her temporary residence permit, but as inGermany, this penalty will rarely be imposed. The failure to fulfil the integrationcontract may also be punished by an administrative fine.In France, the immigration authorities can take the fulfilment of the welcomingcontract into consideration at the first renewal of the residence permit (a draft lawproposes to extend this to any renewal) and in deciding on the application for a per-manent residence permit. Non-fulfilment does not automatically lead to non-renewalof the temporary permit or non-issuance of a permanent permit. The issuance of apermanent permit is conditional on ‘republican integration’ into French society, theevaluation of which leaves wide margins of discretion to the immigration authorities.It is not clear how much weight they attach to the fulfilment of the welcoming con-tract in deciding on applications for permanent permits. However, a video on thewebsite of the OFII suggests that such a permit will not be issued if the welcomingcontract has not been completed. There appear to have been no cases where immi-grants were refused a renewal of their temporary permit on this ground.Unlike the above-mentioned countries, Belgium has not made access to a perma-nent residence status conditional on the fulfilment of an integration requirement.However, the Flemish Region requires some immigrants to attend an integrationprogramme. The failure to comply with this obligation can lead to financial penalties.It may be punished by an administrative fine or by the withdrawal or reduction ofsocial benefits.

Consequences regarding the residence rightsIn most Member States, the rights of migrants are strengthened with the obtainmentof a permanent residence status. A permanent residence permit offers its holdergreater security of residence than a temporary permit. The level of security offeredby the permanent status, differs from country to country. In the Member Stateswhere the permanent residence permit offers the highest security, the permit can belost on three grounds only: fraud, long absence from the country, or long prison sen-tences Third country nationals holding a permanent status in these Member States donot have to fear losing their residence rights in the event of, e.g., unemployment, de-pendence on public assistance, offences against public order, or, if they were admit-ted to the country on humanitarian grounds, in the event of an improvement of thesituation in their country of origin. Furthermore, in many Member States there areother rights attached to a permanent status. Like the level of security, these rightsvary between the Member StatesIn Austria, holders of the long-term resident status have unlimited access to thelabour market, whereas others need an employment permit that is only granted if noAustrian or EEA citizen is available to fill the position. Furthermore, equal access tomany social provisions, in particular social security payments, is dependent on apermanent residence status. Federal or provincial school subsidies are only grantedto pupils who have at least one parent with a permanent resident status. In mostAustrian cities, access to council housing is also dependent on being an EEA citizenor holding the status of a long-term resident.In Germany, holders of a permanent residence permit have advantages with re-gard to family reunification, employment, and finding accommodation.In the United Kingdom, the advantages of indefinite leave over the extension oflimited leave will depend on the particular category. All persons with indefiniteleave have the possibility to sponsor a limited range of dependent relatives to comethe United Kingdom. Indefinite leave also gives eligibility to the full range of socialbenefits (except for persons with refugee leave, humanitarian protection or discre-tionary leave, who have access to benefits automatically). Migrants in sponsoredskilled workers categories obtain freedom of employment, while those in familycategories gain an autonomous status. The costs of immigration applications are afurther consideration. The costs of applying for extension of the permit regularly aremuch higher than the costs for an application for indefinite leave.In addition to formal rights, a permanent resident status may also give its holderaccess to, e.g., a mortgage or an employment contract. In most countries, banks willnot grant mortgages to migrants with temporary residence permits, and many em-ployers will be hesitant to offer them an employment contract.

all cases, an apparent or perceived crisis of integration (Michalowski 2007). Twomain concerns are discernible in the debates in most countries. The first is to makepermanent immigrants economically self-supporting, to lower their unemploymentrates and to reduce the costs they incur to the state in the form of welfare expenses.The second concern, which became more important with the post-2001 wave of ter-rorist activities and unrest associated with Muslim communities in the United King-dom, Spain, France, the Netherlands, Denmark, and Germany is to familiarise immi-grants with the history and culture of the country of immigration and to inculcate inthem the values and principles of liberal democracy. There are other, more latentaims and concerns as well, such as the desire to limit access to permanent residenceto ‘well integrated’ immigrants (which, for that matter, is an increasingly manifestaim in Denmark and the Netherlands) or to assure the native population that thegovernment or the mainstream political parties are managing the crisis efficiently (cf.Joppke 2007).

DenmarkAmong the countries studied, the Netherlands and Denmark were the first to intro-duce compulsory integration programmes for newcomers. The Netherlands did so in1998, Denmark in 1999. A few years later, Denmark was also the first country tomake permanent residence conditional upon the passing of an examination. It did soin 2002, when the liberal-conservative government that came into power in 2001adopted a new aliens policy. The policy was based on three pillars, one of which wasthe strengthening of the requirement of being self-supporting. When the integrationprogramme was introduced in 1999, it was compulsory, but the public responsibilityto provide opportunities for immigrants to integrate on an equal footing with othercitizens was stressed. Since then, there has been a shift to emphasising the immi-grant’s responsibility for his/her own integration. This resulted in 2002 in anamendment of the integration legislation to the effect that immigrants had to pass alanguage examination as a precondition for obtaining permanent residence, and ithas since led to far-reaching amendments. In 2010, in the legislative debate about thelatest amendment, it was explicitly stated that one of the aims was to make it moredifficult for immigrants who are ‘not well integrated’ to obtain permanent residence.

draft bill later the same year. In the political debate, the FPÖ presented the introduc-tion of the integration contract as a major political success and as a paradigm shifttowards a more restrictive immigration policy. The head of the parliamentary factionof the FPÖ described the contract as a device for selection and as a remedy for immi-gration into the welfare system. The SPÖ (social-democrats) and the Greens sharplycriticised the bill of law as preventing integration. The law itself as adopted in 2002(Fremdengesetz 2002) defined the aim of the integration contract as: ‘It is aimed atthe acquisition of a basic knowledge of German in order to facilitate participation inthe economic, cultural and societal life in Austria.’

GermanyIn Germany, the idea of obligatory integration courses for immigrants was first putforward by the Süssmuth Commission (2001). Referring to the experiences with thistype of integration measure in the Netherlands and Sweden, the commission pro-posed introducing similar courses in Germany. Though the ‘right to participate’ wasstressed, the possibility of obliging immigrants to participate was an important in-strument from the beginning. It was justified by the importance of the aim to bereached (the promotion of integration) as well as by the argument that women whowere isolated at home could be accessed and brought into German society using thisinstrument. The integration courses were introduced in the Zuwanderungsgesetz2004, which was approved on a cross-party basis. The aim of the courses as stipu-lated in the law is to acquaint foreigners with the way of life in the Federal Republicof Germany so that they can act independently in all areas of daily life. To the extentthat there was debate on the integration courses, it focused mainly on who was topay (the immigrant or the state, and if the latter, the federal government, the states,or the municipalities) and what type of sanctions (positive or negative) should beused. Eventually compromises were reached on both issues. Although there was re-luctance to emphasise the compulsory aspect of the courses, this aspect has becomeincreasingly important. Currently, the government plans to introduce integrationcontracts based on the French model. These contracts which, just like the integrationcourses, would apply to new arrivals as well as immigrants who have lived in thecountry for years should help to make integration efforts ‘more binding’.

amination. The explanatory memorandum to the bill which was sent to parliament inSeptember 2005 stated that in order to ‘fully participate in Dutch society’, immigrantsneeded to have knowledge of the Dutch language and to know and accept Dutchnorms and values. The most important issue in the legislative debate on the bill waswhether naturalised citizens born outside the EU/EEA should be included in the tar-get group. They were included in the initial bill. The Council of State advised thatthis was discriminatory. The government revised the bill so that specific categories ofnaturalised citizens would be included (recipients of social benefits, parents of unde-rage children, ministers of religion). After a second negative advice from the Councilof State, the idea to include naturalised citizens was abandoned altogether.

United KingdomIn the United Kingdom, integration requirements were unknown in the field of im-migration law until recent years. This may be explained by several factors. One wasthat many of those who migrated to the United Kingdom in the post-war decadescame from territories that were, or had been, British. Another factor was the gradualacceptance of a form of ethnic pluralism by British policy-makers. In the late 1960sand 1970s, the focus was on ensuring equality of opportunity in social and economiclife, while by the 1980s and 1990s, cultural recognition had increasingly come to thefore. Insistence on language or integration requirements would have conflicted withboth of these versions of pluralism. Recent years, however, have seen language re-quirements rolled out at all stages of the immigration and nationality law system,and the introduction of the ‘Life in the UK’ test for first naturalisation and then inde-finite leave. The background to these developments was that the consensus concern-ing an unqualified version of cultural pluralism had come to an end. The catalyst forchange was a series of riots involving young British Muslim men of South Asian ori-gin in Northern English cities and towns in 2001. Official reports highlighted socialsegregation as a major background factor. The initial government response was setout in a report entitled ‘Building Cohesive Communities’. One of its conclusions wasthat there was a need to promote ‘a uniting identity’ and ‘shared values’ to givepeople ‘a common sense of belonging’. From 2002, social and economic participationhas become more important in the policy discourse. When the extension of the re-quirement to show ‘knowledge of life in the UK’ to indefinite leave applications wasproposed in February 2005, it was part of a policy concerning social and economicparticipation: ‘permanent migrants must be as economically active as possible; put aslittle burden on the state as possible; and be as socially integrated as possible’.

born in Latvia during Soviet occupation and who after 1991 did not qualify for Lat-vian citizenship. When the Long-term Residence Directive was transposed into Lat-vian law in 2006, the debate in the Saeima focused mainly on whether these non-citizens should be entitled to EU citizenship and what requirements they would haveto meet to acquire the LTR status. The leftist parties sharply criticised the bill of lawfor requiring non-citizens to apply for the status and to pass a language examinationand not singling them out as a special group. Otherwise, there has not been muchdebate about the introduction of a language examination for applicants for perma-nent residence as well as for the LTR status. There is a broad consensus that know-ledge of the language should be required. The stated aim of the language require-ment is that it is a device for integration into Latvian society.

FranceAs in other countries, the background to the introduction of integration requirementsfor immigrants seeking permanent residence in France was an apparent failure ofimmigrant integration. After the 2002 elections (in which the extreme right were verysuccessful), a centre-right government came into power. It announced that all newimmigrants would be invited to sign acontract d’accueil et de l’intégration(CAI). TheLoi Sarkozy of November 2003 made the granting of a ten-year residence card de-pendent onl’intégration republicaine,defined in the law as ‘knowledge of the Frenchlanguage and of the principles that constitute the French Republic’, but it did notspecify how such integration was to be determined. With the entry into force of theLoi Sarkozy II of July 2006, the CAI became compulsory. The explanatory memoran-dum to the bill explained that this was to reinforce the route towards ‘republican in-tegration’ and to enable a better evaluation of an immigrant’s integration (Micha-lowski 2007; Joppke 2007; Carrera & Wiesbrock 2009). The move from voluntary tocompulsory courses and to making permanent residence conditional upon the ful-filment of the CAI has not aroused much debate in France. This may be explained bythe circumstance that the proposals were part of larger projects, which containedother more sensitive issues.

to attend a language course. A large majority of the ‘newcomers’ in France have suf-ficient French language skills to be exempted from this part of the welcoming con-tract. In other countries, a much larger proportion of the newcomers are affected bythe requirements, either because a large majority are not proficient in the language ofthe country (Austria, Denmark, Germany, the Netherlands, the Flemish Region inBelgium), or because there are no exemption possibilities for immigrants who are al-ready proficient in the language of the test (United Kingdom). Thus, in Austria in theperiod January 2006-June 2009, about 81,000 immigrants enrolled on a language andintegration course. In Denmark in 2008, 37,833 newcomers followed a Danish lan-guage course; about 60 per cent were obliged to do so. In Germany in 2009, 54,062immigrants were obliged to attend a language and integration course; among themwere 33,474 newcomers, 27,746 recipients of unemployment benefits, and 2,842 per-sons who had lived in Germany for a longer period of time (Integrationsbeauftragte2010: 239). In the Netherlands in the period January 2007-December 2009, 127,000immigrants (49,000 of whom were newcomers) received notice from their municipali-ty that they would have to pass the Dutch language and integration examination;83,000 of them (36,000 newcomers) enrolled on a course. In the United Kingdom inthe period April 2007-June 2010, nearly one million immigrants sat the ‘Life in theUK’ test.The attendance rates of those who are obliged to attend a course or programmeappear to be quite high, at least in countries where data are available. For example, inFrance in 2008, 87 per cent of the newcomers who were required to do so did attend alanguage course. In Germany in 2009, the attendance rate of immigrants who wereobliged to attend a language and integration course was 80 per cent, and it is esti-mated that only 8-10 per cent refuse to attend. In the Flemish Region in Belgium, 64per cent of the immigrants who signed an integration contract in 2007 received an at-testation after having attended at least 80 per cent of the classes; the proportion was68 per cent for those who were obliged to attend as against 56 per cent for those whowere entitled but not obliged to (De Cuyper 2010: 53). In the Netherlands, attendancerates have increased since municipalities can oblige immigrants to enrol on a course.The pass rates for the language and integration tests in use in the different coun-tries appear to vary substantially. For example, in France, 90 per cent of the newco-mers who attended a language course in 2008 passed the (level A1.1) test at the endof the course. In the Netherlands, in December 2009, the pass rate stood at 79 per centfor all those who had taken the (level A2) Dutch language and integration examina-tion since its introduction in January 2007 (26,000 out of 33,000 candidates). The passrate for those who sat the examination for the first time in 2009 was 74 per cent. InDenmark in 2008, 87 per cent of those who enrolled for the examination passed it.The pass rate was highest at level A2 (99 per cent) and lowest at level B2 (81 percent). In Latvia, 74 per cent of those who took the (level A2) language examinationrequired for permanent residence in the years 2007-2009 were successful. Among thecandidates were also, however, persons who took the examination to be admitted toparticular professions. In the United Kingdom, too, 74 per cent of the immigrantswho sat the (level B1) ‘Life in the UK’ test in the period July 2007-June 2010 were suc-

cessful.27In Germany, 46 per cent of those who attended a language and integrationcourse in the period 2005-2008 attained the required level of language proficiency(B1); it has since increased to over 50 per cent. The pass rate for the nationally stan-dardised knowledge of society test, which was introduced in 2009, is much higher(90 per cent).Pass rates in themselves do not say much as long as it is not known what propor-tion of the target group has taken the test. In Austria, a wide gap was found betweenthe number of immigrants who enrolled on a language course in the period January2006-June 2009 (81,000) and the number of those whose fees were refunded after hav-ing passed the examination (7,480). Only a small proportion seemed to have beenprepared to sit the examination upon completion of the course; most seemed to post-pone it. In the Netherlands in the years 2007-2009, 127,000 persons were given noticethat they would have to pass the Dutch language and integration examination withinthe next 3.5 years; those who had fulfilled this requirement by the end of 2009 (24,000persons) were probably not representative of the entire population, but above aver-age, motivated and able. The pass rate for the Dutch examination is therefore likelyto decrease somewhat in the coming years.Pass rates broken down by nationality are available for the United Kingdom’s‘Life in the UK’ test.28The pass rates vary greatly by nationality. The global pass rate,for all nationalities, for the period November 2005-June 2010, stands at 71 per cent.The total pass rate for the 17 states which are designated as ‘majority English-speaking’ in the British immigration system is 86 per cent, as against 70 per cent forthe rest of the world. There is however significant variation among the English-speaking states. What emerges is differentiation between developed and developingcountries. New Zealand, Australia, the USA, Canada, and Ireland have a combinedpass rate of 98 per cent. For the other 12 designated states, all in the Caribbean area,the pass rate is only 70 per cent. There is also significant variation among the non-English-speaking states. Whereas the pass rates for Singapore and Japan stand at 95per cent, various other Asian nationalities have pass rates below 50 per cent. Thisvariation again suggests differentiation by level of development. Another impressionfrom the pass rates by nationality is that there is differentiation by immigration cate-gory. Many of the nationalities with relatively low pass rates have had substantialnumbers of persons granted humanitarian status in Britain over the past decade ormore; examples are Sri Lanka, Angola, Kosovo, Afghanistan, Iraq and Turkey.Pass rates broken down by sex, age group and country of origin are available forthe Netherlands. Three years after the introduction of the language and integrationtest, the pass rate stood at 76 per cent for women, and 83 per cent for men. Youngercandidates are more likely to be successful than older ones. The pass rate stood at 85per cent for candidates younger than 36, and 60 per cent for candidates older than 55.27The data were obtained by the UK research team through a freedom of information request to theUK Border Agency. The data do not distinguish between applicants for indefinite leave and fornaturalisation, and the data are by test attempt rather than individual.The data were obtained by the UK research team through a freedom of information request to theUK Border Agency. The data do not distinguish between applicants for indefinite leave and fornaturalisation, and the data are by test attempt rather than individual.

As in the United Kingdom, the pass rates in the Netherlands also vary considerablyby nationality. The pass rate for Turkish candidates is 63 per cent, the pass rate forPolish candidates 93 per cent. It is 90 per cent for candidates from the former SovietUnion, 85 per cent for candidates from former Yugoslavia and Iraq, 80 per cent forcandidates from Afghanistan and 74 per cent for Moroccan candidates.Comparable data are not available for the other countries studied. However, inDenmark, where language courses are offered at three levels (A2, B1, B2), immigrantsfrom non-western countries are proportionally overrepresented in the lowest-levelcourses. Many students who are enrolled on a level A2 course will not be able to passthe level B1 examination (the level required for a permanent residence permit).In Germany, the first results of a longitudinal study show that a younger age, ahigher educational level, and not having been born in Turkey, Russia or another for-mer Soviet Republic or South or East Asia have a positive effect on the progressachieved during the course (Rother 2009).

Effects on Permanent Residence and Security of ResidenceIn most of the countries studied, it is too early to establish the effects of the integra-tion requirements on the number of applications for and grants of permanent resi-dence.The integration requirements which were introduced in Germany in 2005 applyto newcomers who were admitted to the country from 2005 onwards. As a perma-nent residence permit can be obtained only after five years of residence in Germany,the first cohort falling under the 2005 legislation will not apply for it until 2010. Thus,it is too early to assess the effects of the requirements on the number of applicationsfor, and grants of permanent residence in Germany.Immigrants in Austria have to fulfil an ‘integration contract’. The current con-tract was introduced in 2005 and newcomers are required to fulfil it within their firstfive years of residence in the country. Thus in Austria, too, it is too early to establishan effect on the numbers of granted permanent residence permits. However, an im-migrant’s failure to fulfil the integration contract may also have consequences for therenewal of his/her temporary residence permit. In the years 2006-2009, about 46,000residence permits were issued to immigrants in the potential target group. The num-ber of fulfilled integration contracts was much smaller (9,200). A tentative conclusionfrom these figures is that the coming years (when the five-year period for the firstcohorts of immigrants falling under the 2005 legislation expires) may see a largenumber of legal disputes. These disputes will have to clarify whether and when fami-ly migrants who have failed to fulfil the integration requirements within the requiredtime period can be deported.In the Netherlands, passing an integration examination has been a requirementfor permanent residence only since January 2010. The Dutch statistics for the first sixmonths of 2010 do show a decline in the number of applications for permanent resi-dence in comparison to the previous year. This decline is probably at least partly at-tributable to the integration requirement. However, the fee for these applications wasalso raised (Ministerie van Justitie 2010: 31). Moreover, the drop in the number of64

applications for permanent residence could also be due to more immigrants applyingfor naturalisation right away, without first applying for a permanent resident permit,as the integration examination also gives access to naturalisation. To see this effect inthe numbers for naturalisation takes more time.As from April 2007, all applicants for indefinite leave in the United Kingdomhave been required to demonstrate sufficient language and ‘Life in the UK’ know-ledge. The introduction of the requirements has not led to a decline in the number ofgrants of indefinite leave. However, it cannot be concluded from this that the re-quirements have not had deterrent effects, because higher levels of immigration fromthe late 1990s onwards increased the pool of potential applicants for indefinite leave.A further problem is that, even if a reduced propensity to obtain indefinite leavecould be shown, it could be due to other causes, including in particular the increasedfees for these applications in recent years.In Denmark, the number of applications for permanent residence permits has de-clined sharply since 2006 (from an annual average of 12,000 in 2003-2005, to less than5,500 in 2006-2009). At the same time, the proportion of rejected applications rose(from less than 20 per cent in 2003-2005 to on average 40 per cent in 2006-2009).29Thedrop in the number of applications can be attributed to the seven-year residence re-quirement which was introduced in 2002 (and which applied to immigrants who ap-plied for a residence permit from 1 March 2002 onwards; immigrants who had ap-plied for or had been issued a residence permit before that date were eligible for apermanent permit after three years of residence). The drop in the proportion ofgranted applications is probably at least partly attributable to the requirement to passa language test, which was also introduced in 2002, and sharpened up in 2007. Refu-gees have been affected most severely by the new requirements. Before 2006, lessthan 10 per cent of refugees’ applications for a permanent permit were refused; in2008 and 2009, the refusal rate was more than 50 per cent. The refusal rates for familymigrants increased from less than 30 per cent before 2006, to 47 per cent in 2006, 39per cent in 2008 and 36 per cent in 2009. The number of granted permanent residencepermits for both refugees and family migrants will probably drop further in the com-ing years. As of 26 March 2010, all applicants for permanent residence have to fulfilthe more restrictive requirements which entered into force on that date. Moreover,children can no longer apply for a permanent residence permit; they have to wait un-til they are 18 years old, and some will have to wait longer because they will not beable to fulfil all requirements at once.For Latvia, where applicants for a permanent residence permit have been re-quired to pass a language examination since July 2006, data are available only for theyears from 2006 onwards. A tentative conclusion from these data is that the introduc-tion of the requirement has not resulted in a significant decline in the number ofpermanent permits granted.

Effects on IntegrationIn most of the countries studied, the effects of the requirements on the integrationprocess of the target groups have not yet been systematically evaluated. An officialevaluation published in the Netherlands in 2010 stated that it was too early to assessthe effects of the current integration requirements on the integration process of thetarget group. In Germany, a longitudinal study (in which 4,000 students in languageand integration courses are interviewed at the beginning, at the end, and a year aftercompleting their course) is being conducted but has not yet been completed (Rother2009).An effectiveness measurement of Danish municipalities’ integration policiesfound that refugees and family migrants were finding employment or starting educa-tion more quickly than previously. The proportion of refugees and family migrantswho arrived in 2006 and who got a job or started education after one year haddoubled compared to those who arrived in 2000. However, the study did not ex-amine whether the integration requirements (which had been sharpened up since2002) played a role in the progress. According to the study, the progress was possiblyattributable to more efficient municipal policies and the economic situation (Hansen2009).The effects of the Flemish integration policy were evaluated in 2010. The evalua-tion study used different methods, including linked data files and interviews. The re-sults showed that immigrants who had completed an integration course more oftenhad a job than immigrants of the same age, gender, nationality, educational level,etc., who had not started or completed an integration course. The impact on otherdimensions of integration was rather limited. Those who had completed an integra-tion course did not have more interethnic contacts than those who had not started orcompleted a course. Furthermore, half of those who had completed a course did nothave more than just a basic level of Flemish three or four years later, when they wereinterviewed (Pauwels & Lamberts 2010: 119 ff).

We interviewed immigrants and other actors in six of the countries that require im-migrants to attend an integration course and/or to pass an integration examinationafter their admission to the country.30Most of the interviews were with immigrants: 25 immigrants were interviewedin Austria; 20 in Belgium; 14 in Denmark31; 27 in Germany; 25 in the Netherlands; 16in the United Kingdom (127 interviews in total). The immigrant respondents in theNetherlands were interviewed in test centres; most of them had just taken the exami-3031The Latvian research team also conducted interviews, but these interviews focused exclusively onthe naturalisation examination.These 14 respondents were interviewed about their experiences when seeking permanent resi-dence. The Danish research team conducted 12 more interviews with immigrants who had ap-plied for naturalisation.

nation which gives access to permanent residence as well as naturalisation. All therespondents in the United Kingdom had taken the ‘Life in the UK’ test; no interviewswere conducted with immigrants who attended combined ESOL and citizenshipclasses. All the respondents in Belgium had attended an integration programme; 12of them had been obliged to do so (in Antwerp); 8 others had voluntarily attended(in Brussels).In addition to these interviews with immigrants, interviews were conducted withteachers and language school representatives (7 in Austria; 5 in Belgium; 5 in Den-mark; 7 in Germany; 5 in the Netherlands); public officials (6 in Austria; 5 in Bel-gium; 3 in Denmark; 8 in Germany; 5 in the Netherlands); and immigrant organisa-tions and other NGOs (5 in Austria; 4 in Belgium; 5 in Denmark; 7 in Germany; 5 inthe Netherlands; 5 in the United Kingdom). More information about the interviewscan be found in chapter 1.

Opinions on the RequirementsIn general, the immigrants interviewed thought that a language requirement in someform was fair. They felt that knowledge of the language was a precondition for mak-ing a life in their country of immigration. However, in Denmark, some respondentsfound the required level of language proficiency (B1) unreasonably high, and in theNetherlands, many thought that the requirements should apply only to newcomers.In countries which also have knowledge of society requirements, these were lesswell received than the language requirements. In the United Kingdom, a minority ofthe immigrants interviewed thought that the official ‘Life in the UK’ handbook con-tained information which it was important to know; there was also criticism that theinformation was potentially out of date. Some respondents thought that the ‘Life inthe UK’ test was questionable because it set a standard that British-born personscould not meet. Similar criticisms were voiced by respondents in the Netherlands.They felt that they were asked to subscribe to norms and values which were nottheirs, or that they had to learn things which no Dutch-born person would know. InGermany, on the other hand, opinions were more mixed. The advanced languagelearners among the respondents found the orientation course the most interestingpart of their integration programme.In general, the immigrant respondents had positive opinions about the coursesoffered in Austria, Denmark, Germany, the Netherlands and the Flemish Region inBelgium. Most of them did not seem to mind their compulsory nature. Several res-pondents in Austria stated that they would not have invested that much effort inlearning German if they had not been obliged to do so. On the other hand, about halfof the respondents in Denmark stated that they would have learnt Danish anyway –even if it had not been a requirement for permanent residence. In Belgium, the re-search team’s impression was that the respondents in Brussels were somewhat moresatisfied with the courses they attended voluntarily than immigrants in Antwerpwere with their compulsory courses. In the Netherlands and the Flemish Region inBelgium, not only newcomers but also settled immigrants can be obliged to completean integration programme. In the Netherlands, the settled immigrants among the67

respondents objected to this obligation. They thought that it came too late for them.Some of the respondents in Antwerp (who were themselves newcomers) complainedthat the settled immigrants in their class were not motivated to learn Flemish andwere spoiling it for their classmates.Some of the impressions gained in the interviews with immigrants were con-firmed in the interviews with teachers. Particularly in Austria and Denmark, theteachers and language school representatives thought that most of their students at-tended the courses not (just) because they were obliged to do so or because theywanted to qualify for permanent residence, but because they wanted to learn the lan-guage. It should be noted, however, that in both countries (and also in the Nether-lands and Germany) a growing number of students attended the courses voluntarily.Some of the teachers in Germany, on the other hand, thought that the attendancerates for their courses would drop if there was no obligation. One of them pointedout that he had many English-speaking students, who managed to get by quite wellin Germany speaking English.Some of the teachers in Austria and nearly all the teachers and language schoolrepresentatives in Germany and Denmark thought that the required level of lan-guage proficiency (A2 in Austria, B1 in Denmark and Germany) was too high forparticular groups of immigrants. The Austrian teachers also complained that the 300hours of the language module of the integration contract were too short to also im-part knowledge of society to their students. Most Austrian teachers were very criticalof the plans to raise the required level of language competence to B1.According to the local officials interviewed in Austria, quite a few newcomerssaw the language and integration courses as something positive and some were real-ly thankful that the courses were offered. Though some of the officials were critical ofthe compulsory nature of the courses, they also thought that it was extremely helpfulin making women attend the course. The local officials in the Netherlands thoughtthat certain groups of immigrants would not attend an integration course if there wasno obligation. This regards women who are disencouraged by their husband to inte-grate in Dutch society and migrants with a fulltime job and mothers with youngchildren. These two groups are often too busy to prepare voluntarily for the test. Fur-thermore, settled immigrants often objected to having to attend a course and take anexamination. Particularly those who had already attended a course under the pre-vious legislation were indignant that they were not exempted from the new require-ments. Immigrants from the United States and Japan and, more in general, immi-grants who already had a job often reacted with disbelief when informed that theyhad to attend a course and pass an integration examination. In Germany, thereseemed to be regional differences. Whereas local officials in Munich spoke of a ‘run’on the courses in that city, teachers and local officials in Potsdam found it more diffi-cult to attract students. This difference is possibly attributable to regional differencesin employment opportunities for immigrants.

Groups Facing Difficulties Fulfilling the RequirementsIn the United Kingdom, the results from the interviews with immigrants and NGOsconfirmed some of the impressions given by the statistics. The ‘Life in the UK’ test isrelatively straightforward for those with fluent English, but difficult for those with alower level of English, particularly if they have had little formal education, and noexperience with computers. Many in humanitarian categories, some in family catego-ries (especially women) and some older people seem to fare less well.As stated, teachers and language school representatives in Austria, Germany andDenmark thought that the required level of language proficiency (A2 in Austria, B1in Denmark and Germany) was too high for particular groups of immigrants. TheGerman teachers thought that students with little formal education could attain levelA2 if they did their best, but not level B1. According to the Danish language schoolrespondents, there was a group of immigrants who were stuck in the languageschool system because they would never be able to pass the B1 level examination. Be-sides people with little formal education, this group consisted of people who did nothave a talent for languages and people whose first language belonged to anotherlanguage family than the Indo-European. The Austrian teachers pointed out that illi-terate students could not attain level A2 within the 300 hours of language trainingprovided in Austria. This group consisted mainly of women from Turkey and theArab countries with little or no formal education and low self-esteem. The 75 hoursof literacy training were also insufficient to learn to read and write.Various respondents in Denmark pointed out that a large group of immigrantswho had been working and supporting their families and thought that ‘that waswhat was expected of them’ now faced difficulties having to work and take a lan-guage course at the same time. Especially people with little education, refugees andwomen had problems fulfilling both requirements.Teachers and local officials in the Netherlands also thought that particulargroups faced difficulties fulfilling the integration requirements in their country. Theywere thinking of elderly people, illiterates and others with little formal education andtraumatised refugees. Chinese immigrants who had lived in the Netherlands formany years also had difficulties learning Dutch. Immigrants with a job and youngmothers found it difficult to combine fulfilling the integration requirements withtheir normal activities.Various respondents in Austria, Denmark, the Netherlands and the UnitedKingdom pointed out difficulties with getting exemptions for disabled or, e.g., trau-matised persons. Respondents in Denmark thought that the rules lacked transparen-cy. Respondents in the United Kingdom pointed out the absence of a specific proce-dure for exemption on grounds of incapacity before an application was made. Re-spondents in the Netherlands thought that traumatised refugees sometimes had dif-ficulty obtaining the required medical certificate. Respondents in Austria thoughtthat the threshold for exemption on grounds of incapacity was too high.

Effects on IntegrationThe immigrants interviewed had different opinions about the importance of the inte-gration requirements for their integration. In Austria, the majority of those who hadalready passed the examination, found the course helpful in their daily life. Two ofthem reported that it had helped them to get a promotion at their workplace. At thesame time, many stated that in order to feel part of Austrian society, it was not suffi-cient to speak German, there also had to be a tolerant and welcoming climate, whichthey had not found in Vienna. In Belgium, almost all immigrant interviewees re-ported an improved level of integration. Some of them referred to the language skillsthey had acquired; others explained that they were better able to get by without hav-ing to ask for help all the time. A few respondents also reported that they felt lessdiscriminated against after having completed the integration programme. The immi-grant interviewees in the Netherlands also stated that they were better able to man-age in daily life. However, a settled immigrant who had been obliged to attend an in-tegration course after years of residence in the Netherlands stated that she had learnteverything she needed in daily life in practice. In Denmark, half of the respondentsstated that they would have learnt Danish anyway. Also half of the respondentsstated that they had integrated into Danish society not so much due to the languagecourse but to other factors, such as family, friends and work. Several respondentsstressed that they felt excluded by the Danish aliens law; the language course had notchanged this. The immigrant interviewees in Germany stressed that they did not feelpart of German society because of their knowledge of the language, but because oftheir having a job and paying taxes in Germany. In the United Kingdom, nearly allrespondents were sceptical about the likelihood that the ‘Life in the UK’ test wouldlead to integration. One respondent was surprised even by the suggestion that inte-gration was one of the test’s objectives. The high degree of scepticism among the in-terviewees in the United Kingdom is perhaps attributable to the fact that many new-comers in the United Kingdom are already proficient in the language of their countryof immigration. This is different in the other countries studied.Most of the non-immigrant respondents were reluctant to claim that the integra-tion requirements contributed to the integration process of those who have to meetthem. In Austria and the Netherlands, many respondents pointed out that level A2(the required level for permanent residence in both countries) was not sufficient forsuccessful participation in the labour market. The same was said by respondents inGermany about level B1. The respondents did not think that applicants for perma-nent residence should be required to attain a higher level of language proficiency.However, they thought that it was unrealistic to expect an effect on the immigrants’integration in the labour market.In all countries with compulsory courses, there were many respondents whoclaimed that these had emancipatory effects for particular groups, e.g., young moth-ers and (other) immigrants who belong to rather closed communities, and whowould not (be allowed to) attend an integration programme if there was no obliga-tion. Particularly in Germany and Austria, many respondents stressed that the psy-chological effects of the courses were probably more important than the languageprogress made by the immigrants who participated in the courses. They spoke of70

Effects on Permanent Residence and Security of ResidenceIn most of the countries studied, failing to fulfil the integration requirements mayhave consequences for the immigrant’s access to permanent residence. In a few coun-tries, it may even have consequences for the renewal of the immigrant’s temporaryresidence permit.In the interviews in the United Kingdom, several examples were mentioned ofimmigrants who could not apply for indefinite leave to remain because they had not(yet) passed the ‘Life in the UK’ test.The interview results in Denmark confirmed that the sharp decline in the propor-tion of granted applications in Denmark was partly attributable to the requirement topass a language examination. However, many immigrant interviewees stressed thatrather than the examination in itself, the combination with other requirements consti-tuted a problem for them. The (former) requirement of active participation in thelanguage classes was particularly difficult to combine with that of being in full-timeemployment. Some respondents were nervous and confused about the new require-ments which were to be introduced at the time of the interviews. The immigrant res-pondents in Denmark considered it very important to obtain permanent residence.They said that they would therefore continue to make an effort. One respondent de-cided to close his shop to get more time to study for the language examination.The decline in the Dutch numbers of applications for permanent residence sinceJanuary 2010 can be attributable to the integration requirement, but also to more im-migrants applying for naturalisation right away, as the integration examination givesaccess to naturalisation as well as to permanent residence. The interviews in theNetherlands provided some evidence that this might be the case. Most of the respon-dents who were preparing for the examination stated that their final aim was to ap-ply for naturalisation. They wanted to be safe or ‘get rid of all the crap’ connected tonot being a Dutch citizen.In Austria and Germany, local officials and other respondents thought that theincome requirements in these countries constituted a bigger hurdle than the lan-guage and integration requirement. Immigrants with low educational levels wouldhave difficulties fulfilling both requirements. These impressions were confirmed inthe interviews with immigrants in both countries. Several immigrant organisationsreported a growing sense of unease in immigrant communities because of the threat-ening denial of the renewal of the temporary residence permit or the granting of apermanent residence permit. In Vienna alone, more than 3,000 persons who had beenadmitted to the country in 2006 would not yet have fulfilled their integration con-tract.71

The RequirementsSeven of the nine countries in this study (Austria, Denmark, France, Germany, Lat-via, the Netherlands and the United Kingdom) have made access to permanent resi-dence conditional upon the passing of a language (and knowledge of society) test forparticular categories of non-EU/EEA immigrants. In five of the seven countries thereare publicly regulated and/or funded courses to prepare for the test, and in fourcountries (Denmark, France, Germany and the Netherlands), immigrants who belongto the target group are (or can be) obliged to attend these courses.The content of the test differs from country to country, but it always includes alanguage examination. The required level of language proficiency varies from A.1.1in France to B1 in Denmark, Germany and the United Kingdom. The target groupdiffers also from country to country. However, it always includes newcomers whohave been admitted for family reunification, and in most countries, a large majorityof the target group falls into this category. EU citizens and their family members arealways excluded from the target group (although in Denmark they now are invitedto participate, and in several other member States they can participate voluntarily).The family members of own nationals, on the other hand, are always included. InGermany and the Netherlands, not only new arrivals, but also immigrants who havelived in the country for a long time can be obliged to successfully complete a lan-guage and integration course if their language skills are considered to be insufficient.In five countries (Austria, Denmark, Latvia, the Netherlands, the United King-dom), immigrants in the target group are not eligible for permanent or long-termresidence status and the rights attached to it unless they can prove sufficient knowl-edge of the language (and society) of the country. In Germany and France, too, im-migrants seeking permanent residence are in principle required to prove sufficientlanguage knowledge, but the consequences for those who fail to do so are lessstraightforward than in the other five countries. On the other hand, both in Germanyand France as well as in Austria, an immigrant’s failure to comply with integrationrequirements may also have consequences for the renewal of his/her temporary resi-dence permit. However, this penalty will rarely be imposed.In most countries where integration requirements were introduced, there hasbeen a tendency to strengthen them (by raising the required level of language profi-ciency or by sharpening the sanctions for non-compliance) and/or to extend the tar-get group.

Effects on IntegrationIn most of the countries studied, the effects of the requirements on the integrationprocess of the target groups have not yet been systematically evaluated. Most exist-ing evaluation studies have been restricted to the implementation of the require-ments.The immigrants interviewed for this study had different opinions about the impor-tance of the integration requirements for their integration. Most of the other respon-dents (teachers, public officials, NGOs) were reluctant to claim that the integrationrequirements for permanent residence contributed to the integration process of thosewho have to meet them. Many of them thought that the required levels of languageproficiency for permanent residence were not sufficient for successful participationin the labour market. The respondents did not think that applicants for permanentresidence should be required to attain a higher level of language proficiency. How-ever, they thought that it was unrealistic to expect an effect on the immigrants’ inte-gration in the labour market. On the other hand, in all countries with compulsorycourses, there were many respondents who claimed that these had emancipatory ef-fects for particular groups, e.g., young mothers and (other) immigrants who belongto rather closed communities, and who would not (be allowed to) attend a course ifthere was no obligation.

Effects on Permanent Residence and Security of ResidenceIn most of the countries studied, it is too early to establish the effects of the integra-tion requirements on the numbers of applications for and grants of permanent resi-dence. In Denmark, however, the requirement to pass a language test appears tohave caused a decline in the proportion of granted applications for permanent resi-dence. Refugees have been affected most severely. The results of the interviews inDenmark confirmed that the drop in the proportion of granted applications waspartly attributable to the requirement to pass a language examination. However,many immigrant interviewees stressed that rather than the examination in itself, thecombination with other requirements constituted a problem for them. The (former)requirement of active participation in the language classes was particularly difficultto combine with that of being in full-time employment.

In Latvia, the requirement to pass a language test does not appear to have led toa decline in the number of permanent permits granted. Likewise, in the United King-dom, the requirement to demonstrate sufficient language and ‘Life in the UK’ know-ledge has not led to a decline in the number of grants of indefinite leave to remain.However, it cannot be concluded from this that the requirement has not had deter-rent effects, because the pool of potential applicants for indefinite leave has increasedin the past decade. Interviews in the United Kingdom confirmed that in a number ofcases migrants were not able not apply for indefinite leave to remain because theyhad not (yet) passed the ‘Life in the UK’ test.In the other countries studied, it is too early to establish an effect on permanentresidence. In Austria and Germany, most respondents thought that the income re-quirements in these countries constituted a bigger hurdle than the language and in-tegration requirement. However, in both countries, an immigrant’s failure to complywith the language and integration requirements may also have consequences for therenewal of his/her temporary residence permit. In the coming years, legal disputesmay have to clarify whether and when this penalty can be imposed on family mi-grants.With regard to the consequences of not obtaining a permanent residence permit,it can be concluded that, apart from not gaining security of residence, in a number ofcountries this can lead to less access to the labour market, housing and social bene-fits. Those practical consequences of being unable to exercise a permanent residenceright as well as the more insecure position can hinder integration. In this way inte-gration requirements can have a counterproductive effect on the integration of immi-grants who face difficulties fulfilling the requirements.

Groups Facing Difficulties Fulfilling the RequirementsThe (future) effects on permanent residence are likely to be more severe for somegroups than others. With regard to the ‘Life in the UK’ test, data on pass rates for dif-ferent nationalities as well as the results from the interviews indicated that many mi-grants in humanitarian categories, some in family categories (especially women) andsome older people therefore fare less well. In Denmark, data on applications for andgrants of permanent residence indicated that refugees have been affected most se-verely, and this was confirmed in the interviews with language school representa-tives. In Germany, the first results of the above-mentioned longitudinal study indi-cated that a higher age, a lower level of education and having been born in Turkey, aformer Soviet Republic or South or East Asia had a negative effect on the progressachieved during a German language and integration course. Most of the teachers andlanguage school representatives interviewed in Denmark, Germany and Austriathought that the required level of language proficiency (B1 in Denmark and Ger-many, A2 in Austria) was too high for immigrants with little formal education.Teachers and local officials in the Netherlands also thought that elderly people, illit-erates and others with little formal education and traumatised refugees faced diffi-culties fulfilling the integration requirements in their country.

Currently, integration tests are applied as a condition for naturalisation in seven ofthe nine countries under consideration in this study. In Belgium, immigrants apply-ing for naturalisation are not required to prove integration, since the condition toshow ‘sufficient willingness to integrate’ was abolished with the introduction ofchanges in Belgian nationality law in 2000 (De Hart & Van Oers 2006: 325). In Franceapplicants for naturalisation are required to have assimilated into the French com-munity, which means that they are required to have sufficient knowledge of theFrench language and the rights and duties conferred by French nationality. Thisknowledge is however not tested in a test, but in an interview with a civil servant atthe local préfecture. Applicants are required to be able to speak and understandFrench in order to be able to face daily and basic communication. In the interview,the civil servant can take the personal circumstances of the applicant, such as his/hereducational background, social condition and age, into account.32In the other countries under consideration, knowledge of the language and/orknowledge of the society is either tested in an integration test, or proof of such know-ledge, for instance by submitting a secondary school diploma, is required. Prior tothe introduction of these requirements, language skills and integration were often as-sumed (UK), or tested in an informal interview (Denmark, Netherlands, some Ger-man states). The requirements are shown in table 4.1 and will be briefly described be-low:

of the applicant could be taken into account. The language requirement was streng-thened on 1 March 2006. Since that date, knowledge of the German language at levelA2, both orally and in writing, has been required. Language skills at level A2 are alsorequired in order to obtain a permanent residence permit. Hence those who fulfilledthe language requirement at permanent residence stage are not required to fulfil thelanguage requirement when applying for Austrian nationality. Furthermore, a natu-ralisation test, in which applicants for naturalisation have to prove ‘basic knowledge’of the democratic order of the Austrian Republic, the history of Austria and the his-tory of the province of residence, was introduced on 1 March 2006. The test is a mul-tiple- choice test consisting of 18 questions, which are drawn from a set of publishedsample questions. The curriculum of the test follows the curriculum of the fourthyear of compulsory secondary school.

DenmarkDenmark requires its future citizens to have language skills at level B2 of the CEF,which is the highest level of language skills required in the countries examined inthis study. Language skills at level B2 have been required since 2006. Prior to 2006and starting in 2002, language skills at level B1 were required, whereas before 2002,applicants for Danish nationality were merely required to be able to speak and un-derstand the Danish language. In 2008, the pass mark threshold for the Danish lan-guage test was increased. A multiple-choice knowledge of society test was intro-duced in Denmark in May 2007 (infødsretsprøve). The test, which costs about €89and can be taken twice a year, consists of 40 questions, five of which concern recentevents, and the remaining 35 are based on a textbook, which e.g. deals with Danishhistory from the beginning of the Viking age (750 a.d.), geography, population, lan-guage, immigration, democracy, literature, art and science. In November 2008, thenumber of questions that were needed to be answered correctly in order to pass thetest was raised, the time offered to complete the test was shortened and the questionswere no longer accessible in advance; thereafter the pass rate in the test dropped (seebelow). In summer 2010, after the content of the knowledge of society test had beencriticised by experts in history, political science and other branches of science and itturned out that for some questions none of the multiple-choice answers were correct,the Minister of Integration announced that she would consider replacing the mul-tiple-choice test with a test consisting of open questions.

evidence, for instance in the form of a language certificate, that they have commandof the German language orally and in writing at level B1. Level B1 equals the level oflanguage which is tested in the final test of the integration courses within the frame-work of the Residence Act (Aufenthaltsgesetz), as is the case in Denmark for level B2language courses, the certificate immigrants obtain when they pass the final test atthe end of the integration course can hence be used to fulfil the language requirementfor naturalisation. Furthermore, since 1 September 2008, applicants for naturalisationhave been required to show sufficient knowledge of the legal system, the society andthe living conditions in the Federal Republic of Germany (Article 10(1)(6) of theGerman Nationality Act) by passing the so-called ‘Einbürgerungstest’ (naturalisationtest). The naturalisation test consists of 33 multiple-choice questions, of which at least17 need to be answered correctly. Whereas the content of the naturalisation test isbased on the content of the integration courses (Article 10(7) German NationalityAct), passing the orientation test at the end of the integration course of the ResidenceAct will not suffice for naturalisation.In the German state of Baden-Württemberg, a number of specific questions areadded to the integration test. If there are doubts that the applicant has carried out ac-tivities which are not based on the Constitution and an inquiry to the Office for theProtection of the Constitution supports these doubts, then the applicant must dispelthese doubts in an interview with the naturalisation authority.

HungaryIn Hungary, applicants for naturalisation are required to pass an examination onconstitutional issues. Knowledge of the Hungarian language is not tested as such, buta high level of Hungarian is required to fulfil the other requirements for naturalisa-tion. The constitutional examination is taken orally and in writing, and only thosewho master Hungarian perfectly will generally pass it. The examination currentlyconsists of 21 topics defined in a textbook, which mentions the knowledge of pupilsin the final year of elementary school as the threshold for the knowledge tested.However, according to the Hungarian rapporteur, the textbook is based on theknowledge of a moderately well educated person with good linguistic competence.The topics include Hungary’s constitutional structure and state symbols. In 2006, thetopics of the examination were extended to literature, art, science and national histo-ry, starting from the establishment of the state in the 10th century. This thematic ex-tension of the examination was an internal decision by the Ministry of Interior andthe amendment was passed as a technical requirement. The content of the test wascriticised after a weekly journal found that 90 per cent of Hungarians interviewed inthe streets could not answer the questions.The Hungarian constitutional examination is managed by local administrative of-fices (20 in total), which organise the examinations before a board generally consist-ing of lawyers, administrators and teachers (the examination board). Whereas theformat of the examination and its major components are regulated by GovernmentDecree, each county administrative office has its own in-house rules on the manage-ment and procedure of the examination including how to assess the oral and written80

LatviaThe naturalisation requirements in Latvia were adopted in 1994 after a profound po-litical debate. Future citizens are required to have written and oral command of Lat-vian at level B1. Knowledge of society is also tested. The language examination con-sists of two parts. The first part tests the applicant’s ability to listen and understand,to read, write and communicate on topics of daily life. This part takes 90 minutes.The examination consists of multiple-choice questions. The writing part consists oftwo tasks: completing a form with personal information and writing an essay. Thesecond part of the examination is a 15-minute interview. A person is allowed threeattempts to pass the examination. If these attempts are not successful the naturalisa-tion application has to be made anew.Applicants who take the knowledge test orally have to demonstrate their know-ledge of the national anthem and answer questions. Thirty minutes are granted forpreparation and the examination itself takes 15 minutes. Answering the questions inwriting takes 45 minutes. In this case, knowledge of the national anthem can be dem-onstrated orally or in writing. The knowledge examination consists of both multiple-choice and open questions. Ten general questions on the history of Latvia and eightquestions on the Constitution are included in the examination.

quired since July 2004. Applicants can also fulfil the language requirement by show-ing sufficient knowledge of Welsh or Scottish Gaelic. Prior to 2004 language skillswere also required, but the language requirement was of limited effect, since therewas no procedure specified by which language capacity could be assessed. Home Of-fice policy was to assume that the language requirement was met in case of ‘circums-tantial evidence’ such as education in British establishments or employment in a Brit-ish firm. Enquiries would only be made in case of indications that the language re-quirement would not be met. On 1 November 2005, the requirement to have ‘Life inthe UK’ knowledge was introduced. Since that date, both knowledge requirements,i.e. knowledge of the language and knowledge of Life in the UK, can be fulfilled bypassing the ‘Life in the UK’ test at level B1. An alternative route to citizenship ishowever offered to those with language skills below level B1: they can follow an‘ESOL with citizenship course’, in which they need to make progress to the next levelof language proficiency. There are hence two ‘routes’ to citizenship. As in the Nether-lands, the same integration requirements apply for permanent residence.34Thosewho fulfilled the knowledge requirements by either passing the Life in the UK test orby submitting a certificate of attendance on an ESOL course for ILR will hence notface new requirements when applying for citizenship.A proposal that fulfilment of an ‘active citizenship’ requirement will shorten theperiod of residence required for citizenship has been made. The fate of the proposalat the time of writing was uncertain.

All of the countries under consideration offer some form of preparation for the lan-guage and/or integration test for naturalisation. Whereas this statement may appearself-evident, in the Netherlands no possibilities for preparation existed for the natu-ralisation test, the predecessor of the integration examination, which applied until 1April 2007. Furthermore, the content of the naturalisation test was undisclosed andno sample questions were published. Currently, candidates for the integration ex-amination can follow an integration course, in which knowledge of the language andof Dutch society are taught, or practise the different parts in the examination on theInternet. In most cases, taking part in the preparation courses will be funded by themunicipalities.35Courses are also set up by the administrative offices in Hungary. The costs ofthese courses vary between €5-10. Candidates can however also choose to followcourses set up by NGOs or private agencies, or study from a textbook. The text of thebook was criticised by an agency providing preparatory courses because of the dryand old-fashioned style of language.

No state-regulated preparation courses for the knowledge of society test are of-fered in Austria, Denmark, Germany, Latvia and the UK. From the interviews it be-came apparent that such courses had been set up by volunteers in the UK. In Austria,the largest provider of courses for the integration agreement had set up a course in2006, but no one had enrolled. Furthermore, the association had never been contactedby immigrants wishing to find a course. Hence, apparently there was no need felt byimmigrants to prepare for the test on a course. Latvia organised courses during thenineties, with external assistance. Courses were supposed to be set up in Germany,but in the end this hardly occurred because of limited demand. One-day preparationcourses are however offered by the adult education institute (Volkshochschule) ofMunich and in Hamburg, politics classes for immigrants, offered since the 1970s, arecurrently used as a means to prepare the topics of the test. These classes are offeredin the form of a ten-day seminar for which educational leave can be obtained. Insteadof attending a course, people can prepare for the knowledge of society test using theInternet, since all questions and answers are published. Furthermore, booklets havebeen published. To prepare for the language test at level B1, immigrants can followan integration course within the framework of the Residence Act. Costs for thecourse are normally €645, but reduced fees may apply.36In Austria and Denmark, immigrants can also prepare for the language require-ment on a language course. In Austria, following module 2 of the integration agree-ment, which consists of 100 to 600 hours of language teaching, will lead to level A2.Costs for the course vary between €1,500 to €2,500.37In Denmark, free three-year lan-guage courses are offered to foreigners who are covered by the Integration Act.However, only about 53 per cent of the students are enrolled on a language course atlevel B2 – the level required for naturalisation.As regards the knowledge of society test, in Austria candidates can prepare forthe test by studying the preparation material published by the federal and provincialgovernments, the published sample questions, or by a training programme whichcan be bought at the website www.staatsbuergerschaft.com. The training programmecosts €12 for the basic version and €18 for the advanced version. In spring 2010 thelearning brochure on democratic structures and history of Austria was removed fromthe website of the Interior Minister following critical questions by the Green Party. In2008, the brochure had been criticised in a scientific article in the journal for Germanas a Foreign Language for inter alia containing factual mistakes and didacticallyquestionable explanations. Furthermore, according to the authors of the article, ex-plaining Austrian history at the level of the fourth grade of secondary school usinglevel A2 of language competence was not possible. In February 2010, the brochurewas again criticised by a renowned political scientist who stated that the brochurewould by no means realise the aim of the Nationality Law to inform about the demo-cratic structure of Austria at the level of the fourth grade and would rank below anacceptable level of a scientifically sound and didactically meaningful civic education.3637Taking part in the course is free for recipients of social welfare and unemployment benefits, seeparagraph 3.3 above for more detailed information.50 per cent of the costs may be refunded if the test is passed within a certain amount of time, seeparagraph 3.3.

Whereas the Interior Minister has stated that the brochure would be evaluated, re-written and republished on the website, this so far has not occurred. Candidates forthe knowledge of society test will hence have to consult other sources to prepare forthe test.Danish candidates in the infødsretsprøve can prepare by studying a textbookwhich can either be bought, viewed on the Ministry of Integration website, or down-loaded as an MP3 file. Until November 2008, all questions which could come up dur-ing the test were published on the Internet, together with the answers. However, thissystem was criticised (and the pass rate of almost 100 per cent was considered toohigh), and in November 2008 it was decided to no longer publish the answers on theInternet.38Applicants for naturalisation in Latvia can prepare themselves by studying abook on the Latvian language examination (€2.60), a book on basic questions on Lat-vian history and Constitution (€ 4.80) and a book with recommendations on the me-thods of preparation for the examination on Satversme, the national anthem and his-tory (€2). The lack of courses is a major problem in Latvia. Projects on integration doexist, but are short-term. Recently the number of providers for courses increased andcourses are offered for €35. However these types of providers just offer their partici-pants exactly the same material as the Naturalisation Board distributes for free.These free EU-supported programmes are offered to target groups, such as parents ofchildren belonging to minorities and teachers. However, the number of participantsis rather low because of limited resources. The Agency recently released an onlinelearning programme at level A1, which is available in both Russian and English.Finally in the UK, candidates for the Life in the UK test can prepare by studyingfive of nine chapters from the second edition of the handbook ‘Life in the UK: AJourney to Citizenship’ (£9.99), published by the Stationery Office, which is the offi-cial publisher, but a private company. The Stationery Office also published prepara-tion material, namely a study guide and a practice q&a with 400 sample questionsand answers based on the handbook. It should be noted that, whereas the advisorycommission on the implementation of the knowledge requirement advised the publi-cation of a handbook for those with a reasonable command of English, they never in-tended the handbook to become the basis for a standard test on knowledge of Life inthe UK. Rather, the commission envisaged that the requirements should be metthrough some form of education, also for those competent in English. The use of thehandbook and the testing of knowledge in the Life in the UK test hence show that thesystem has departed far from the educational intentions of the advisory group.

Table 4.2: CostsAustriaA training plan for the test costs €12 (basic version) or €18 (advanced version).The preparation script and the naturalisation test are free of charge. Dependingon the state where the naturalisation takes place, and the title under which thenaturalisation is granted, fees range from €621 to €1,878 . Costs for the lan-guage course vary between €1,500 to €2,500.Not applicable.The costs for the citizenship test are €89. A fee of €134 has to be made uponapplication. Renewed applications are free of charge.Not applicable.Language test costs €100 or higher, except Berlin where the costs are €23. TheEinbürgerungstest costs €25. A fee of €645 has to be paid for the languagecourse. Furthermore, a fee of €255 has to be paid for the application.The fee for the language test is €20.Handbooks with prices varying between €2 and €4.80. Courses are offered for€35 (but they do not add anything to the free courses). Free courses are also of-fered by the State Language Agency. The fee for naturalisation is €28. A fee of€4.30 is set for the poor and unemployed, family members of families includ-ing three or more minor children, pensioners, severely and moderately dis-abled persons and persons attending duly accredited education establish-ments. Persons exempted from fees are the politically repressed, very severelydisabled, orphans and children who are not cared for by parents and personswho are registered for social benefits.Immigrants who are offered a course by their municipality can be required tocontribute €270 to the costs of their course and exams. For others, the costs aremuch higher. The fees for the three parts of the central exam are €126 in total.The fee for the practical exam is €104 (portfolio) or €250 to €1,200 (assessment).Those who choose to follow a course will also have to pay a course fee. Thecosts are partly reimbursed if the integration exam is passed within 3.5 years. Itis also possible to get a loan.The fee for the short exemption test is €81 and the fee for the NT2 state exam€90.Prices for a single application are €567. In case someone jointly applies withhis/her partner the fee is €719. Stateless persons and those who are granted aresidence permit on asylum grounds receive a €50 reduction on each applica-tion.The fee for the ‘Life in the UK’ test is £34. The official ‘Life in the UK’ hand-book costs £9.99 in hard copy. The costs of the ESOL courses (for immigrantswhose level of English is below B1) vary depending on the provider and thekind of course. There are exemptions from fees for, e.g., persons in receipt ofsocial benefits and persons in humanitarian categories, and reduced fees forpersons protected by EU free movement law and the partners of settled per-sons who have been resident in the United Kingdom for one year.The fee for a single application is £780 and in case someone jointly applies withhis/her partner the fee is £1010.

All immigrants wanting to obtain citizenship via naturalisation are required to fulfilthe language and knowledge of society requirements for naturalisation. The conse-quences of not fulfilling the requirements will be that the nationality for the countryunder consideration will not be granted, and that the immigrant will retain an infe-rior legal position, be it either as a legal permanent or temporary resident. Generallyspeaking, two categories of persons can however be exempt from passing the neces-sary tests: those whose knowledge of the language and the country is already consi-dered to be sufficient, and those of whom it cannot be expected that they will ever beable to fulfil the requirements due to mental or physical impediments or age.

4.3.1 Exemptions on the Basis of Sufficient IntegrationNative speakersIn Austria, native speakers of German are exempt from submitting a language certifi-cate when applying for naturalisation. In the Netherlands, immigrants from Belgium(Flanders) or Surinam can be exempt if they can submit a high school diploma prov-ing that a pass mark was obtained for the subject ‘Dutch’.Applicants following education or having obtained a diplomaApart from the UK, in all of the countries under consideration, immigrants who havefollowed education in the country or are still in school, are exempt from showingproof of sufficient language skills and/or passing the knowledge of society test as acondition for naturalisation. They are considered to have sufficient language skillsand/or knowledge of society.The UK does not provide an exemption ground for im-migrants who are either in school, or who have obtained a diploma. All applicantsfor naturalisation will have to submit a Life in the UK test certificate (or an ESOL-with-citizenship course certificate) when handing in their applications.Apart from in the UK, the language and knowledge of society requirements willhence mostly apply to first generation immigrants, who have not gone to school inthe country in which they want to naturalise, and to school dropouts. Furthermore,in some countries the proof of having passed a certain examination (if recognised tothat end) is considered adequate proof of the necessary language skills or exemptionfrom the language and/or knowledge of society requirements (Austria, Denmark).Therefore, besides first generation migrants, the integration requirements mainly al-so apply to those second generation immigrants with lower intellectual capacities.Lastly in Hungary, the exemption regulation for immigrants with a diploma mainlyseems to have been introduced to the benefit of ethnic Hungarians. Immigrants whohave completed their secondary schooling in the ‘host’ country will generally besecond generation immigrants, who in most of the countries under consideration canbenefit from facilitated access to citizenship such as option or declaration.In Austria, having a secondary school diploma will not lead to exemption fromeither the language requirement or the knowledge of society test. As regards the lan-guage requirement, only those who are in school can be exempted from proving suf-86

ficient language skills (level A2). Furthermore, exemption is made conditional on ob-taining a positive grade in the subject ‘German’. Exemption from the knowledge ofsociety test can only be acquired if someone has obtained a positive grading in thesubject of ‘history and civics’ at the level of the fourth year of compulsory secondaryschool. Those who have come to Austria past the compulsory schooling age and whofor instance have obtained a diploma granting access to University or a Universitydegree, will hence not be exempt from either the language requirement or the know-ledge of society test.In Denmark, the entitlement to naturalisation for second generation immigrants,hence those born, raised and gone to school in Denmark, was repealed in 2004. Thus,second generation migrants need to fulfil all naturalisation requirements. Most ofthem will, however, be able to prove their adequate skills in the Danish language byhaving passed grade 9 or 10 of Danish public school with a mark of 4 or higher ineach Danish discipline (a mark of 2 is sufficient to pass the school leaving examina-tion and a mark of 7 is considered an average). All applicants for naturalisation willhave to pass the knowledge of society test. The interviews with police officers inDenmark showed that the majority of applicants for naturalisation, i.e. between 70and 80 per cent, prove Danish language skills by submitting a diploma from a Da-nish school. This statement suggests that most applicants for naturalisation aresecond generation immigrants who attended school in Denmark and who until 2004had an entitlement to citizenship. This may indicate that many first generation immi-grants face difficulties fulfilling the naturalisation criteria, among others the highlanguage requirement.In Latvia persons who graduated from elementary, secondary or higher educa-tion where the instruction language was not Latvian, and who have passed a centra-lised examination in Latvian, are exempted from the language test if they apply with-in two years following this examination. Applicants who have been educated in Lat-vian are not bound by such a time limit. Both groups have to take and pass theknowledge examination.In the Netherlands, secondary school diplomas or higher will lead to exemptionfrom passing the integration examination. Furthermore, persons who are not follow-ing secondary education or who have not (yet) obtained a secondary school diplomaare exempt if they have followed education in the Netherlands for a period of eightyears during the obligatory schooling age (5-16). This exemption clause has appliedsince 1 January 2007. Before, school dropouts or those still in high school would notbe exempt from passing the test. However, in practice it appears that civil servantsapply the list of diplomas leading to exemption very strictly. A recent example is thatof a Danish applicant with a PhD in Dutch language and literature who had to takethe language test for naturalisation, because this study was not on the list of the civilservant.Germany, like the Netherlands, exempts immigrants who have obtained a Ger-man secondary school diploma or higher or who have attended a German schoolfrom proving sufficient language skills. Four years ‘successful’ attendance in a Ger-man speaking school suffices for exemption. Furthermore, those who progress to thetenth class of the intermediate secondary school, comprehensive school or grammarschool are exempt from the language requirement. To be exempt from passing the87

Einbürgerungstest, only certificates of ‘general’ education lead to exemption. Anacademic degree will hence not necessarily lead to exemption from the knowledge ofsociety requirement. In Baden-Wurttemberg this ruling has been interpreted asmeaning that only those who obtained a university degree in public administration,law and politics are exempt from the naturalisation test.In Hungary, students studying for a certificate in Hungary are exempt from pass-ing the constitutional examination. Since 2006, persons who attended public elemen-tary school where the subjects were taught in Hungarian, whether in Hungary or inanother country, are also exempt. The introduction of this exemption clause means afacilitation of the naturalisation procedure for ethnic Hungarians, the majority ofwhom will have attended public school with curricula in Hungarian. A visit to thewebsite of the Office for Immigration and Nationality Affairs shows that problemshowever often occur regarding the burden of proof. A further extension of the ex-emption regulation to the benefit of ethnic Hungarians is foreseen. Starting on 1January 2011, immigrants with a Hungarian ascendant or whose origin from Hun-gary is probable will be exempt from passing the examination if they can prove theyhave knowledge of Hungarian.In all of the countries mentioned above, only proof of sufficient integration ‘inblack and white’, i.e. on the basis of a generally rather limited list of diplomas, willlead to exemption from fulfilling the language and knowledge of society require-ments. Persons who have integrated in a different way, for instance through workingand residing in the country of which they want to become full members, will be re-quired to submit proof of integration by submitting language and knowledge of so-ciety test certificates. In the Netherlands, complaints uttered by municipal officialswho saw themselves obliged to refer immigrants, who were obviously well inte-grated, to an integration course or an integration test, led to the preparation of anamendment of the exemption regulation to provide for a possibility to exempt obvi-ously integrated immigrants.

4.3.2 Exemption on the Basis of Age or an ImpedimentAgePersons who have reached a certain age will be exempt from fulfilling the languageand knowledge of society requirements in Austria, France, Germany, Hungary andthe UK. The exact age someone will need to have attained is not specified in Austria.Persons who want to be exempt from fulfilling the requirements on age grounds willneed to submit a certificate from a public health official that they are unable to fulfilthe requirements due to their health. In Germany, persons who have reached the ageof 60 and who have been living in Germany for a period of 12 years will be exemptfrom the requirements. In Hungary and the UK, persons aged 65 or over will be ex-empt.A general exemption on age grounds is not provided in the Netherlands and inDenmark. In Denmark, until 2002, persons aged over 65 were exempt from showinglanguage skills in the integration interview. With the formalisation of the languagerequirement, this exemption possibility for older immigrants was repealed. In the88

Netherlands, those who have reached the age of 65 can opt for Dutch citizenship,which means that they are not required to pass the integration examination. Optionis however only possible after a period of residence of 15 years.In Latvia there is also an exemption for persons below a certain age. Naturalisedpersons can opt for citizenship for their children under 15 years old. However, inpractice a problem of so-called ‘forgotten children’ has arisen, i.e. where parents ac-quire citizenship but forget to file an application for registration of their children.When these children reach the age of majority they have to follow the normal proce-dure for naturalisation.Mental or physical impedimentsIn all countries under consideration, exemption from fulfilling the language andknowledge of society requirements may be obtained if someone suffers from a men-tal or physical impediment. The impediment needs to be proven in a doctor’s attesta-tion in which the relationship between the impediment and the inability to fulfil therequirements is made clear.In Denmark, possibilities for exemption on the basis of a mental or physical im-pediment are rather limited, following a reduction of these possibilities in 2006. Ex-emption is only granted under exceptional circumstances in cases where the appli-cant is suffering from a very serious illness. Immigrants suffering from PTSD (post-traumatic stress disorder) are explicitly excluded from obtaining exemption. Deci-sions regarding exemption are taken by the Naturalisation Committee behind closeddoors and in principle no reason for refusing the application for exemption is given.The immigrant organisations interviewed stated that being granted exemption wasalmost impossible and some stressed that the rules were rather unclear. Several re-spondents had never experienced a case where a person had been exempted; one or-ganisation even saw blind people’s applications for exemption turned down. Anumber of the respondents called for more grounds for exemption, especially as re-gards PTSD.In the UK also the exemption procedure appears to be applied with rigidity. Inthe UK, staff members from law centres complained about the absence of a specificprocedure for exemption before an application is made. Those who want to be ex-empt need to file their applications together with the documents on the basis ofwhich exemption might be granted. If exemption is not granted, the applicant willlose the application fee, even if further information or documents might have led tothe application succeeding. One law centre once saw the application of a deaf anddumb person refused. Only after issuing a pre-action protocol saying that judicial re-view would be taken if the decision was not reviewed, was the application granted.IlliteracyIn most of the countries under consideration, illiteracy is not explicitly mentioned asan exemption ground. Only the Netherlands and the German state of Bavaria ap-peared to have introduced a special procedure for illiterates. In the German state ofBavaria, a special language test, the so-calledAlpha-test,is used for illiterates. Thereis, however, no adapted knowledge of society test for illiterates. Since reading and

writing skills are required to complete this test, passing it will prove an insurmount-able barrier for illiterates.In the Netherlands, those who are illiterate or disabled have the possibility of ex-emption from the naturalisation test. An illiterate candidate who wishes to qualifyfor exemption has to be illiterate in both Dutch and his or her first language, and hasto show that a serious attempt has been made to learn Dutch. Subsequently, illiteratecandidates have to undergo a feasibility investigation They will be exempted fromthe naturalisaton test only if the outcome of this feasibility investigation is that theywill not be able to learn Dutch at A2 level within the next five years. Since 1 April2007, the immigrant is however in this case also required to pass the ‘test spokenDutch’. For the feasibility investigation, €287 is charged. The fact that illiterate candi-dates have to undergo an expensive examination, for which they have to travel toAmsterdam, before they can be exempted from the naturalisation test or, since 1April 2007, the integration examination, may constitute a barrier to naturalisation forthis category of immigrants.

Changing opinion on the function of naturalisationAs we have seen above, prior to the introduction of formal language and knowledgeof society tests, basic oral language skills, which were tested in an informal interview,were required for naturalisation in Austria, Denmark, some German states, and theNetherlands. At that time, naturalisation in Denmark and the Netherlands was re-garded as a step (Netherlands) or a ‘crucial positive element’ (Denmark) in the inte-gration process.In the Netherlands, the the Minorities’ Policy departed from the central idea thata good legal position would be beneficial to an immigrant’s integration. This viewformed the basis for the requirements for naturalization, as codified in the 1985Rijkswet op het Nederlanderschap (Dutch Nationality Act). Naturalisation became aright rather than a favour. In order to further open up access to Dutch nationality, therenunciation requirement was abolished in 1992. Christian Democrats in the Parlia-ment in the course of the 1990s however started to express a different opinion on theplace of naturalisation in the integration process than the government, of which theyalso formed part, had done so far: instead of a step in the integration process, natura-lisation should be the end of a completed integration. Furthermore, the rise in thenumber of naturalisations, which was especially high after the abolition of the renun-ciation requirement, according to the Christian Democrats showed that naturalisa-tion had become ‘too easy’. They therefore started lobbying for the introduction ofwritten language skills and knowledge of society as conditions for naturalisation.Eventually, these requirements were introduced after sufficient support was gained90

from other political groups. After the introduction of the naturalisation test, the Con-servative Liberal Minister for Integration repeatedly referred to naturalisation as ‘thecrown’ on the integration process.In Denmark, the view that naturalisation is a means rather than an end of inte-gration was rejected by the centre-right Danish People’s Party (DPP). In 2002, theDPP entered into an agreement about the naturalisation criteria with the governingparties (Conservatives and Liberals), after which inter alia the language requirementwas strengthened. Hence, the governing parties, as the DPP had always done, alsorejected the idea of naturalisation as a means of integration. Instead, they contendedthat ‘the acquisition of citizenship presumes that the applicant is already integratedin Danish society’.39Whether naturalisation was explicitly seen as a means for integration by the Aus-trian government prior to the introduction of the knowledge of society and languageskills at level A2 requirements in 2006 is unclear. However, when the Austrian right-wing Freedom Party (FPÖ) entered into a coalition government with the ChristianDemocratic ÖVP in 1999, it expressed the view that naturalisation should come at theend of the integration process. Previously, criticism regarding the naturalisation pro-cedure expressed by the FPÖ had already led to the introduction of the condition of‘adequate knowledge of the German language’ in 1998. Like the Christian Democratsin the Netherlands, the FPÖ argued that acquisition of Austrian nationality was ‘tooeasy’. The introduction of a language requirement apparently did not have the effectintended by the FPÖ, since it started a stronger lobby for stricter naturalisation crite-ria in 2005. The high number of naturalisations according to them showed an unfa-vourably liberal attitude to nationality. The new rules regarding naturalisation cameinto effect on 1 March 2006.40Hence, both in the Netherlands and in Austria, the wishto reduce the number of naturalisations explicitly played a role in the debates preced-ing the reinforcement of the language and integration criteria for naturalisation.

Increased integrationIn many of the countries under consideration, the introduction and strengthening ofthe language requirements and the introduction of the condition of societal knowl-edge have been justified using the argument of improved immigrant integration. Theway in which this desire was worded however varied.In Austria, for example, naturalisation is seen as a culmination of integration,which can only be reached with a good knowledge of German. In Denmark, lan-guage and societal requirements must secure that new Danish citizens have suchknowledge of the Danish language, culture, etc. that they can get along in Danish so-ciety, follow the developments and debates and thus participate in the Danish society3940Letter from the former Minister for Integration dated 7 June 2006.Besides the requirement to prove language skills at level A2 and the introduction of the know-ledge of society test, other requirements for naturalisation were also strengthened. The residencerequirement was increased and naturalisation is refused to applicants who have received socialsupport payments in the three years preceding the applications.

on an equal footing with the rest of the Danish population. Social and economic par-ticipation in society was also primarily the focus of policy justification in the UK. Ex-amination for citizenship was perceived to strengthen the ability of new citizens toparticipate in society and to engage actively in the democracy.In Germany, in the eyes of the government, the formalisation of the language re-quirement was necessary to guarantee the immigrants’ participation in the politicaldecision making process. Only the possession of oral language skills, which was con-sidered to be sufficient in some states, was considered insufficient for proper politicaldecision making skills.In Hungary, the examination of constitutional issues was introduced with theexpectation that applicants should have fundamental historical and constitutionalknowledge by passing an examination demonstrating their attachment to Hungary.The test was also considered to contribute to the moral relevance of Hungarian citi-zenship.In Latvia, naturalisation requirements, especially the language requirement, wereconsidered to be a basis for building a civil society and for defining a shared systemof values. During the Soviet occupation many immigrants from the former Soviet re-publics came to Latvia. After regaining independence, Latvia had to ensure state con-tinuity. This implied that fundamental laws and the Constitution in force before theSoviet occupation had to be revived and only the persons who were citizens at thetime of the occupation as well as their descendants were recognised as Latvian citi-zens. In order to ensure the existence of Latvia, including its own identity, the lan-guage requirement was compulsory, while many persons were bilingual (Latvianand Russian) or had no knowledge of Latvian at all. The naturalisation requirementsset by the Citizenship Law in 1994 were hence aimed at persons who came to Latviaduring the Soviet era.

A more uniform interpretation of the language requirement for naturalisationIn Germany and the Netherlands, the wish for a more uniform interpretation andapplication of the language requirement led to a formalisation of the language test. Inthe Netherlands, this formalisation at the same time implied a strengthening of thelanguage requirement. In Germany, it meant a strengthening of the requirement intheLänderthat previously only tested oral language proficiency.Research conducted in the Netherlands in 1988 had pointed out that the lan-guage requirement was applied differently in the municipalities. Whereas the ma-nual for the application of the Dutch Nationality Act stated that only oral languageskills were required, ten per cent of the municipalities also required written languageskills. The 1992 bill, which proposed the official abolition of the renunciation re-quirement, hence also contained a reformulation of the language and integration re-quirement. A more precise formulation of the language and integration requirementwould enhance equality and fairness in its application.In Germany, next to a different application of the language requirement, the re-quirement to declare loyalty to Germany’s free and democratic basic order also led todifferent practices in the Länder. It was hence decided to add the requirement of92

knowledge of the legal and social system and the way of life in the Federal territory,generally to be proven in a naturalisation test, as a condition for naturalisation. Sinceonly the state of Baden-Württemberg employed questions within the framework ofthe declaration of loyalty, the introduction of the naturalisation test implied astrengthening of the overall naturalisation procedure in the other German states.

4.5 Consequences of not passing the test

Migrants who are not exempted from the requirement to pass the test, will not beable to naturalise as long as they do not pass the test. This has consequences for theirpolitical rights, especially the right to vote or to be voted as a politician. Although insome Member States migrants enjoy voting rights on the municipal level if they fulfillcertain conditions (for instance after a legal residence of five years), participation inregional or national elections is reserved for citizens. Furthermore, certain profes-sions can only be practised by citizens, like the profession of judges or certain admin-istrators. Which professions belong to this category, varies from country to country.Another consequence of not obtaining citizenship is that migrants do not enjoy thehighest level of security of residence. Although the exact conditions for withdrawalvaries from country to country, they are in all Member States more strict regardingcitizenship than regarding the permanent residence permit. These three kinds of con-sequences (political rights, professional rights and security of residence) can influ-ence the degree of integration of migrants.

Number of applications, naturalisations and refusalsIn Austria, Denmark, Latvia and the Netherlands the number of persons who natura-lise is declining. In Germany differences between the federal states exist, since thefederal Nationality Act is implemented differently in the states. In France the lan-guage and other requirements do not appear to have affected the number of applica-tions. Hence, the number of naturalisations have remained more or less similar overthe years. For Hungary there are no statistics on this issue. Because of the differencebetween the countries, each country will be described individually below.In Austria there has been a continuous decline in the number of naturalisationssince 2004, in particular since 2006. The number of persons was 26,259 in 2006 and14,041 in 2007 with further declines in the follow-ing years. This decline can be ex-41Unless otherwise stated, the figures in this section have been taken from the national INTEC re-ports.

plained by several factors. The large group of refugees from former Yugoslavia be-came smaller because they had reached the ten years of residence required for natu-ralisation between 1998 and 2004. Another reason can be found in the restriction ofnaturalisation criteria. According to some of the interview respondents, elderly mi-grants who speak poor German may be put off the idea of naturalisation because ofthe test. Most respondents thought, however, that the stricter residence and incomerequirements are more likely to explain the decline in naturalisations. Furthermore,Directive 2003/109/EC was implemented in Austrian law, which equalised the statusof long-term residents to that of Austrian citizens in terms of access to the labourmarket, social rights and access to housing. This took away some of the importanceto naturalise.In Denmark the number of naturalisations decreased after 2002 due to implemen-tation of stricter rules in that year. To a large extent the drop in 2003 was caused bythe formalisation of the language requirement, which meant that applicants wereforced to put their applications on hold until they had passed a language examina-tion at level B1. In 2004 and 2005 the number increased again because more and moreimmigrants had taken the necessary language examinations. However, with the riseof the required language level from 2006 (from B1 to B2) a new decrease has occurredand this time the decrease appears to have a more permanent character. Further-more, the number of refusals of applications increased in 2002, 2003 as well as in theperiod 2007-2009. Refusals have increasingly been given due to a lack of Danish lan-guage proficiency. However, at the moment two new important grounds for refusalaccount for the largest percentage of the number of refusals: the requirement forpassing the naturalisation test and the requirement for self-support.In Germany, there was a decline in naturalisations from 2000 to 2006, while in2006 the numbers increased slightly. However, in 2007 and 2008 numbers decreasedagain, followed by a slight increase in 2009. At the same time, the number of foreign-ers eligible for naturalisation on the basis of their time spent in Germany is high. In2007, only 2.9 per cent of those eligible on such grounds were naturalised (Integra-tionsbeauftragte 2010: 443). Mention should be made of the fact that naturalisationrates differ considerably between the German states and cities because of different in-terpretations of the naturalisation requirements, differences in the length of the pro-cedure and differences in administrative quality. The number of naturalisations in-creased in seven states, whereas there was a decrease in nine states. The organisationof a campaign to promote naturalisation has proven to have a positive effect on thenumber of naturalisations in Stuttgart, where the number of applications rose by tenper cent in 2009.In Latvia the abolition of the so-called age window system in 1998 and the acces-sion of Latvia to the EU in 2004 led to a considerable rise in the number of naturalisa-tions. However, after 2006 the numbers dropped significantly, probably becausesince then non-citizens could also apply for the longterm residence status if theywanted to get a certain level of freedom of movement within the EUsThe level of thetest to require this status is lower than the test regarding Latvian citizenship.In the Netherlands a differentiation can be made between the period before theentry into force of the revised Dutch Nationality Act in 2003, introducing the forma-lised language and integration requirement in the form of the naturalisation test and94

reinforcing the residence requirement, and the period after this revision.42The num-ber of applications for naturalisation was much higher before the coming into forceof the revised Dutch Nationality Act. Compared to 2002, the year before the revisedDutch Nationality Act entered into force, the number of applications for naturalisa-tion decreased by two-thirds in 2004. In 2008, the number of applications were stillmore than half as low as the number of applications filed in 2002. From 1997 to 2001,hence prior to the introduction of the revised Dutch Nationality Act, a decreasingtrend in the number of applications can also be seen.43The decreases in the numberof naturalisations in the years prior to the introduction of the revised Dutch National-ity Act however were not as sharp as the decrease which can be noticed after. Theconclusion can hence be drawn that the revision of the nationality legislation hasbeen the most important cause of the strong decrease in the number of applicationsfor naturalisation. In 2007, the naturalisation test in the Netherlands was replaced bythe integration examination. Other requirements for naturalisation remained unal-tered. The new decrease in the number of naturalisations in 2007 and 2008 can henceprobably be ascribed to the introduction of the integration examination, which hasthe same level as the naturalisation test, but which is more extensive and expensive.In the years to come, a rise in the number of naturalisations can however be ex-pected, since there has been a rise in the number of persons passing the integrationexamination (see below).In the UK, the number of naturalisations have not declined since the formalisa-tion of the language requirement and the introduction of the knowledge of ‘Life inthe UK’ requirement in the years 2004 and 2005. On the contrary: one can even see anincrease in naturalisations despite the strengthening of the requirements. In the fourcalendar years prior to 2004, the average number of naturalisations was 78,564 and inthe four full years since 2005, the average number was 117,441. An explanation forthis is the fact that higher levels of migration to the UK from the 1990s onwards in-creased the pool of potential applicants for both ILR and naturalisation. This upwardtrend makes it difficult to identify any deterrent effect since the language and Life inthe UK requirements were introduced. Another problem is that even if a reducedpropensity to obtain naturalisation could be shown, this could also be due to othercauses, such as the increased fees for the application.44

Pass ratesPass rates for the knowledge of society tests in Austria and Germany are high. InAustria according to an article in the weekly ‘News’ of 1 August 2007, the pass ratefor the test at first attempt was on average 90 per cent in Austria. Three other reportson the success rate for naturalisation testing show higher pass rates, e.g. 95 per centin 2010. In Germany from the day theEinbürgerungstestentered into force (1 Septem-ber 2008) to 31 March 2010, 106,831 people took part in the test, with a success rate of98.5 per cent. Pass rates also used to be high in Denmark, where until December 200897 per cent of all candidates passed the test successfully. After the introduction of anew test in December 2008, the pass rate dropped considerably: more than three-quarters of the candidates failed the test. Since then the pass rate has risen, being 54.5per cent in December 2009, and there is reason to believe that the pass rate for June2010 will further increase. It should be borne in mind, however, that as well as theknowledge of society test, applicants for naturalisation in Austria, Denmark andGermany will also have to fulfil the language requirement which especially in Den-mark appears to form a high barrier (see analysis of interviews in paragraph 4.6).In Latvia, the pass rates for the naturalisation examinations have been decliningover the years. During 2004 and 2005, when naturalisation reached its peak, the per-centage of those who failed the knowledge test was 3.7 per cent and 4.8 per cent re-spectively, but by 2007 and 2008 those figures had already reached 10.4 per cent and17.8 per cent. The results for the language examination are similar. During 2004 and2005, respectively 10 and 16 per cent of all participants in the language examinationfailed. In the years thereafter, the failure rates increased, reaching a peak in 2009when almost 40 per cent of all language examination candidates failed. This can beexplained by two factors: the decrease in the offer of courses financed by the stateand the lower educational level of the current group of applicants.The Hungarian report mentions an estimated success rate of 70 per cent in theconstitutional examination. From interviews it becomes clear that the failure rate is10 per cent. Furthermore, the second attempt is in most cases successful. The compo-sition of the Examination Board has a strong influence on the pass rate.In the Netherlands the pass rate for the naturalisation test, in force from 1 April2003 to 1 April 2007, was 60 per cent of all those who enrolled for the test and even-tually passed it. The others did not turn up after they were confronted with the obli-gation to pay for the test (as well as the fee for naturalisation), or they did not turn upfor the second part of the test or failed the test. Departing from the number of actualparticipants, the pass rate for the test is higher, namely 74 per cent. On 1 April 2007,the integration examination replaced the naturalisation test as a condition for natura-lisation. Since the integration examination is also a condition for permanent resi-dence, the results in this examination are considered in paragraph 3.3 of this report.With a pass rate lying between 72 and 74 per cent, the pass rate in the integration ex-amination is more or less similar to the pass rate for the naturalisation test. The abso-lute number of persons passing the integration examination is however much higher.Whereas a total of 14,300 persons passed the naturalisation test during the four yearsit was in force, with 49,500 persons passing the integration examination in the period2007-June 2010, the number of successful candidates in this test appears to be three96

Number of exemptionsStatistical data on the number of immigrants who are exempted from the languageand knowledge of society requirements are only available for Denmark and theNetherlands. In Denmark the number of exemptions from the language requirementhas decreased, possibly because the practice of the Naturalisation Committee hasbeen restricted in parallel with the restrictions on the requirements for naturalisation.In the Netherlands in the first one and a half years after the introduction of thenaturalisation test, around 85 per cent of all applicants for naturalisation were ex-empt from passing the test on the basis of a diploma. Around three per cent of allapplicants were exempt due to language or medical impediments, which leavestwelve per cent of all applicants who actually took the test (INS 2004: 33). More re-cent statistics show that in the years 2005 to 2008, 4.5 per cent of all the applicantswhose application was granted was exempt on the basis of a medical or languageimpediment. Furthermore, the percentage of persons who successfully passed the na-turalisation test before applying for naturalisation has risen to between 25 and 29 percent (INS 2007: 71).45The majority of all applicants for naturalisation is hence exemptfrom passing the required test, mostly on the basis of a diploma.

AgeIn the countries that had data as regards the age of the naturalised immigrant (Ger-many and the Netherlands) the age group of persons between 18 and 35 accounts formost of the naturalisations. In Germany naturalised migrants are increasingly be-coming younger. The average age of persons who were naturalised in 2007 was 30.5years and the average age of persons who were naturalised in 2009 was 29.5 years. InLatvia approximately one-third of the naturalised non-citizens consists of personsfrom 18-30 years and the number of older people opting for naturalisation is decreas-ing. By representing two-thirds of the candidates who passed the naturalisation test,the group of immigrants aged 18 to 35 are by far the most represented age categorythat registered for and passed the test in the Netherlands as well. Representing one-third of the total, immigrants aged between 35 and 65 represent the second largestgroup who registered for and passed the naturalisation test. Immigrants aged 65years or older are the least represented category among immigrants registering forand passing the test. Furthermore, the pass rate for this age category is significantlylower than the pass rates for the younger age categories. With 62.2 per cent, the pass45Since accurate data on the number of persons passing the test before applying for naturalisationwere not available, the percentage was based on an estimate using available numbers.

EthnicityApplications for naturalisation and/or pass rates for the tests broken down by natio-nality are available for Hungary, Germany, Latvia, the Netherlands and the UK. InHungary 90 per cent of naturalised persons are native ethnic Hungarians. A spotcheck among 101 applications in 2009 showed that 69 per cent of the applicationswere made by ethnic Hungarians. In Germany most naturalised foreigners are fromTurkey (24,647 in 2009). In Latvia the largest share of naturalised citizens are of Rus-sian ethnicity who have submitted 68 per cent of all naturalisation applications in the14 years since the introduction of naturalisation. This corresponds to their shareamong the non-citizens of Latvia (on 1 July 2009 66 per cent of non-citizens wereRussian). The following groups are Belarusians (10 per cent applications and shareamong non-citizens 13 per cent) and Ukrainians (9 per cent of applications, 9.5 percent among the non-citizens).In the Netherlands and the UK, pass rates in the tests broken down by nationali-ty are available. In both countries, the pass rates clearly differ between the countriesof origin. In the UK, it appears that pass rates for countries with a high level of de-velopment are higher. The pass rates furthermore suggest that there are differencesof outcome per immigration category: countries from which many refugees andfamily migrants originate generally have low pass rates (see above paragraph 3.7).In the Netherlands the pass rates also vary considerably between the nationali-ties.46As in the UK, the introduction of the test has resulted in a selection of futurecitizens in which the country of origin plays an important role.

In general, examinations testing language and societal knowledge are supported bythe interviewees in all countries. Both immigrants interviewed in Austria who werepreparing for the naturalisation test approved the duty to take a test and saw it as aself-evident right of the state to impose such an obligation. A similar feeling of un-derstanding as regards the obligation to meet naturalisation requirements exists inDenmark and Germany, where in most cases naturalisation tests (and orientationcourses) are considered a good opportunity to learn about history and current affairsand to gain an understanding of society. Furthermore, the Naturalisation Office andthe police in Denmark indicate that migrants who currently apply for naturalisationhave a better knowledge of the Danish language than former applicants. However,fewer immigrants apply for naturalisation, which is probably due to the language re-quirements. In Latvia citizens and non-citizens generally support the examinations.Latvian officials and language teachers are positive about the naturalisation process.Although a general sense of understanding is present among interviewees, criti-cal comments have been made. The immigrants interviewed in Denmark can be di-vided into two groups, low educated migrants to respectively higher educated mi-grants and/or second generation migrants. Half of the applicants for naturalisationinterviewed were born and/or raised in Denmark. Some of the respondents of thislatter category think it is ‘unreasonable’ or ‘ unfair’ that they have to comply with na-turalisation requirements. The second generation migrants consider themselves Da-nish and feel offended that they have to prove this by taking a test. They can, howev-er, prove their language skills by showing a diploma from a Danish public school. Itappears that in Denmark the well-educated immigrants form the largest part of natu-ralised immigrants. In Latvia some non-citizens felt humiliated that they had to takea test. Two NGOs supported this view. However, survey data show that the numberof persons who think that the naturalisation process is humiliating has decreased.Most academic graduates in Germany believe that the test is not necessary for thembecause they are already equipped with an appropriate level of relevant knowledge.Other migrants showed little understanding for the meaning of the naturalisationtests: ‘I live and work here. What a German does, I am also doing. Why should I notvote and why must I take the test?’

knowledge is tested in an interview. Nevertheless, in general migrants consider thelanguage requirement as the most difficult part of the test.In Germany and Latvia the naturalisation test is not perceived to be difficult. InGermany, it was questioned, both by migrant advisory services and migrants,whether the naturalisation test makes an individual into a loyal citizen and an equalmember of society. Naturalised respondents in Latvia indicated that the tests consistof simple questions and that the contents are relevant. Conversely, non-citizens thinkthat the examinations should be easier, especially for the elderly. All Latvian officialsand NGOs interviewed noted that the most difficult part of the examination was thewritten part of the language examination. Furthermore, the Office of Citizenship andMigration Affairs received complaints about history questions, such as ‘what hap-pened on 17 June 1940?’.In Denmark persons in all categories interviewed indicated that the naturalisa-tion requirements, especially the level of the language requirement, were too high.According to immigrant organisations the language requirement is the hardest to ful-fil and it is criticised, also by language teachers, for excluding less well educated im-migrants from becoming Danish citizens. Some organisations point out as a goodthing that the conditions are standardised. Some migrants, furthermore, stated thatthe content of the knowledge of society test is, for example, irrelevant or superficial.With regard to questions on history there are strong doubts whether young autoch-tonous Danish citizens would know the answers. Language teachers in Denmarkcomplained that what was tested was if ‘the applicant has learned the text book byheart’.In the Netherlands the answers by the migrants show a wide variety of whatthey think of the test. Their opinion on the usefulness of the test seems to be moreuniform. In particular the ability to speak the language is seen as a very importantaspect to function in everyday life. Part of the critique by the different interviewees isdirected at the content of the knowledge of Dutch society test: are these questionsuseful and do autochthons know the answers? Some immigrants considered thequestions regarding behavior and values and norms as unnecessary brainwashing.Furthermore, municipalities indicated that compiling the portfolio places a heavyburden on the migrant.In the UK a common assessment indicates that the handbook contains unneces-sary information and potentially outdated information. Test candidates confirm thatthe test is relatively straightforward for those with fluent English language skills.One respondent said that the test asked things British citizens did not even know.Some candidates had ideas for reformation of the tests, such as an interview to assessthe candidate’s Life in the UK knowledge. Also two candidates thought the test wastoo easy for what it purported to assess.Although the examination in Hungary only constitutes of a knowledge test onconstitutional issues, the questions are so difficult that a high level of Hungarian isrequired (at least C1 level). Officials therefore only advise immigrants with high lan-guage abilities to take the examination. This explains the low failure rate despite thefact that the test is very difficult.

Preparation for the testIn all countries respondents prepared themselves for the test, for example, by follow-ing a course, self-study or experience in practice. In Germany most test candidatesprepared for the test on the Internet. Orientation courses contain complex vocabularyand require a higher level of proficiency than A2/B1 levels. Immigrants who possesslimited German language skills often pass naturalisation tests by learning the an-swers by heart.In Austria, according to teachers, candidates only rarely attend a languagecourse. In most cases naturalisation takes place after at least ten years of residence inthe country and it can be expected that persons have acquired enough knowledge ofthe German language by that time to pass the language test at Level A2. Some candi-dates are, however advised to attend a course if the official in charge believes that theapplicant has insufficient knowledge of the language. Teachers furthermore indicatethat there is no demand for language courses and that immigrants prepare their testat home. Two immigrants interviewed were indeed advised to follow a course beforetaking the test and they were well informed about the possibility of downloadinglearning material.In Denmark the language teachers stated that there were not enough possibilitiesfor preparation for the knowledge test. In the Netherlands a number of the candi-dates for the naturalisation test prepare for the test by following an integrationcourse. This course has been introduced within the framework of the Integration Act.As prior to the entry into force of this act, candidates for the naturalisation test werenot supported in their preparation for the test, it can be concluded that the Integra-tion Act (which also includes the integration requirement for a permanent residencepermit) facilitates the preparation for naturalisation. The suggestion was made byteachers in the Netherlands that more attention could be paid to ordinary day-to-dayworries or cultural affairs. Most migrants were satisfied with the contents of thecourses. Some officials of the municipalities however mentioned the fact that a lot ofattention was paid to passing the tests (teaching for tests) and that the main goal –learning the Dutch language – might disappear out of sight.As regards preparation in the UK all test candidates relied on the handbook, sev-eral also on questions and answers and many had visited unofficial Internet websitesto test their skills.The major fear of examinees in Hungary is that they will not understand thequestions asked at the examination, because they differ from the preparatory materi-al. Fear for the examination appeared to be a major reason for Latvian non-citizensnot to apply for Latvian citizenship. A number of them felt too old and had insuffi-cient knowledge of the language. Due to the lack of Latvian courses, this group is notable to overcome its insufficient knowledge.

Motives to naturaliseAccording to Austrian NGOs naturalisation is more and more becoming an expres-sion of the decision to stay in Austria forever. This was confirmed by one of the mi-grants interviewed. Naturalisation would also prevent immigrants from legal dis-crimination and practical hassles. One immigrant wanted his children to becomeAustrian citizens and therefore the whole family decided to naturalise. According toNGOs the interest has dropped since the introduction of the long-term residence sta-tus, in particular among older migrants, who often feel too old to learn the Germanlanguage well enough and to pass the test. This group is also less interested in freemovement and voting rights. It appears, according to NGOs, that young people alsooften opt for the EU long-term residence status instead of naturalisation, as this sta-tus is easier to obtain and facilitates movement within the EU.In Germany the reasons given for naturalisation were predominantly of a prag-matic nature: to have the freedom to travel, to have a German passport and avoidstress when dealing with the authorities, to join dependants who had already becomeGerman. In some cases, the right to political participation was also mentioned as amotive. The ability to travel was also one of the main motives to naturalise in Latvia.Other main motives, emotional as well as pragmatic, were for example the wish tobelong to the country, have voting rights, the fact that the applicant would continueto live in Latvia, the partner was a citizen of Latvia or the fact that citizenship wasneeded in order to find a job. Some non-citizens did not opt for naturalisation be-cause they wanted to maintain closer ties with Russia. Loss of their status would bean obstacle for travelling to Russia.Most respondents in the Netherlands were participating in a course and/or thetest because they were under the obligation to integrate, but indicated that after thetest they wanted to apply for naturalisation. This integration obligation of the Inte-gration Act made the barrier to opting for naturalisation lower. The most importantmotive particularly for naturalisation was that these migrants wanted to be safe, or,to put it differently, to ‘get rid of all the crap’ connected with not being Dutch. Inter-est groups, particularly refugee organisations, also mentioned this consideration, in-dicating that naturalisation was the ultimate guarantee for refugees of not runningthe risk of being sent back to their country of origin. Migrants themselves mentionedmainly practical reasons: improvement of their current situation, or as a requirementfor further education or a (better) job. They also indicated that the possession ulti-mately, of a Dutch passport would grant them the most secure position. Also the factthat some of these migrants had partners who already had a Dutch passport stimu-lated them to get in an equal position. A particular practical motive, as in Germanyand Latvia, was the idea that the possession of a Dutch passport would allow theholder to travel far more easily around the world.

test in Austria. Also in Latvia older persons face more difficulties with the languageexamination than others and are often afraid of taking it. However, younger peoplehave more problems with the knowledge examination.In Hungary having Hungarian language skills is the highest barrier for non-Hungarian speaking (non-ethnic) migrants. Non-Hungarian speakers are screenedout through this examination, which means a hindrance in the acquisition of natio-nality. According to respondents the Latin alphabet and Hungarian pronunciationare difficult for Russian, Serbian, Chinese or other Asian examinees.In the Netherlands, certain groups find it difficult to meet the requirements: el-derly people and low educated migrants, illiterates in particular, Asian migrants andtraumatised refugees.In the UK the test is relatively straightforward for those with a good command ofEnglish, but language and Life in the UK requirements have greater impact on cer-tain groups. Many persons in humanitarian categories, some in family categories –especially if they are women – and some older people fare less well. Those groupsare now less free to naturalise than they would have been had the naturalisation re-quirements not applied.

ExemptionsIn four countries, the results of the interviews contain some information on the opin-ions as regards exemptions. One respondent in Denmark finds the possibilities forexemption too poor. A number of immigrant organisations call attention to the needfor more grounds for exemption, especially as regards post-traumatic stress disorder.Applications for exemption from the language requirements are often submitted onphysical or mental health grounds or on word-blindness. The Danish NaturalisationOffice notes that only very serious health problems that prevent the person from ful-filling the requirements constitute a valid ground for exemption. Latvian NGOs anda British legal advisor argued for more exemptions on grounds of ages. British re-spondents mentioned the absence of a specific procedure for exemption on groundsof incapacity before an application is made. In the Netherlands, a release of the re-quirement for traumatised migrants can be given on medical grounds. However, notall doctors are familiar with traumata of refugees, which, according to teachers andrefugee interest groups, result in unreal demands and unnecessary stress.In the Netherlands, Germany and Austria respondents complained about thelimited number of officially recognised diplomas that creates an exemption from thisrequirement. Dutch municipalities and teachers mention cases in which certain di-plomas do not appear on the list when they should. In Germany, the way authoritiesdeal with language requirements is a problem. In one case, the authority in charge re-fused to recognise a language certificate that had been issued on an earlier date, al-though the naturalisation applicant spoke fluent German (8. Lagebericht 2010: 445).In the Netherlands and Austria, an academic degree from a university in the countrydoes not count as fulfilment of the language and societal knowledge requirements. Inthe Netherlands, a Danish assistant professor in Dutch literature had to do a shortexemption test in order to prove her language skills. In Austria, a Mexican citizen103

Problems and obstacles experiencedIn Denmark it was pointed out as a problem that the test is held only twice a year,since candidates will have forgotten the knowledge they gained in that time.In Hungary the requirements for the level of Hungarian as well as the level ofknowledge have not been defined. The missing transparency of the examination or-der and assessment methods has attracted critics. In the Netherlands similar prob-lems occur, as the content of the test is kept secret, although there is some exercisematerial available. Not knowing what will be asked increases the stress of migrantstaking the test. Municipalities and course institutions also mention the lack of feed-back to the candidate as regards the mistakes he/she has made as a downside of thelearning effect of a candidate.In the UK the potential difficulties, according to a volunteer who ran studyclasses, include a lack of education in the country of origin or experience of comput-ers. Immigrants are concerned and scared that the process is complex and they arenot used to studying and taking examinations. Another problem indicated by a mi-grant organisation and a law centre was the lack of time for preparation due to workobligations. A law centre also mentioned the situation where it was difficult to getfamily support to get a woman out of the house and into a language course.In the Netherlands, the level of the short exemption test is B1, while the level ofthe naturalisation test is A2. Despite a national judgment that this higher require-ment is not in compliance with the Integration Act, the authorities did not lower thelevel.

Effects on integrationIn some countries, doubts were expressed on the presumption that a naturalisationtest would lead to a better integration in society. For the majority of respondents inDenmark, passing the test on knowledge of society did not lead to a feeling of beingintegrated into Danish society. In Germany and the Netherlands, immigrants whohad no problems meeting the requirements already felt part of society and stated notto be better integrated after meeting the requirements. The test resulted in the feeling,especially of higher educated migrants, that their integration was unappreciated. Ac-cording to some German officials the tests are not an appropriate method to inte-grate, because someone could pass the test without understanding or dealing withthe content. Furthermore, the certified level B1 is not sufficient for any profession,because the language ability at this level is quite elementary. This is why immigrantswho pass the tests can improve their economic situation only to a limited extent.Dutch language teachers also criticised that the actual ‘certificate’ (A2) had no valuewhatsoever regarding subsequent training or education; it could not be comparedwith the so-called basic qualification needed for further education or the job market.104

In Latvia immigrants appreciate naturalisation differently from language teach-ters. Immigrants do not consider naturalisation as a means to facilitate integration.Non-citizens (especially ethnic Russians) do not feel invited to acquire citizenship,mainly because of historical reasons and personal feelings regarding Latvian citizen-ship. Most of them thought that the test had an adverse effect on integration. Con-versely, language teachers say that examinations are necessary for integration intosociety. Language and knowledge of history are crucial to making daily life easierand for making persons feel better at work and positive about Latvia.In the UK there was general agreement among candidates that the test did notcontribute to their integration. A frequent statement that was made was that the testwas only about memorisation, and was therefore not suited to test integration. Ac-cording to two respondents of advice centres, minority groups seek the company ofpersons from the same background, e.g. a Somali would socialise with other Somalis,whether they spoke English or not. One volunteer at a migrant organisation howeverpointed to the positive effect of tests: they would give the person a sense of achieve-ment and would make them feel more confident.

The requirementsSeven of the nine countries under consideration in this study (Austria, Denmark,Germany, Hungary, Latvia, the Netherlands and the UK) apply integration tests as acondition for naturalisation. The knowledge of language and a knowledge of societyare tested, or proved in another way, for instance by a relevant diploma. All of theabove mentioned countries offer some form of preparation for the language and/orintegration test for naturalisation. In four of the seven countries (Austria, Denmark,Germany and the Netherlands) there are publicly regulated and/or funded courses toprepare for the language test. In the Netherlands and Denmark, these courses werenot introduced for the preparation of the naturalisation test, but within the frame-work of the Integration Act. As the test on the basis of this act has the same contentas the naturalisation test, the courses serve in practice as a proper preparation fornaturalisation. No state-regulated preparation courses for the knowledge of societytest are offered in Austria, Denmark, Germany, Latvia and the UK. Hungary is theonly country where knowledge of language is not explicitly tested. In France the lan-guage requirement is tested in an informal interview and in Belgium no language re-quirement is applied.In the six countries where knowledge of language is required the level differs.The level is the highest in Denmark (B2). As regards the knowledge of society test itis difficult to estimate in which country this requirement is the hardest to fulfil.It appears that once tests are introduced, requirements for naturalisation haveonly been strengthened. In Austria, for example, where a language requirement fornaturalisation has been applied since 1998, the requirement was made uniform andstrengthened in March 2006. In the same year also a knowledge of society test was in-troduced. This trend of adding knowledge of society requirements to the language105

test in naturalisation procedures, can also be seen in Denmark, Germany, the Nether-lands and the United Kingdom. The strengthening of the requirements affects thenumber of applications, naturalisations and refusals. The decrease in these numbersis partly due to restriction of naturalisation criteria, but is also influenced by otherfactors. In the United Kingdom for example an increase can be observed, despite thestrengthening of the requirements. This is presumably related to the increase in thenumber of immigrants in the last decade.In the Netherlands and the United Kingdom the same integration requirementsapply for naturalisation as for permanent residence. Hence, no new requirements areapplicable when a person applies for citizenship. It is thus to be expected that thenumber of naturalisations in the Netherlands will increase, since there has been a risein the number of persons passing the integration examination. With regard to ‘old-comers’ , this effect could be less evident, as for them the required level for the grant-ing of permanent residence is lower. In Austria and Germany fulfilling the languagerequirement for permanent residence means exemption from this requirement whenapplying for naturalisation.Although in general testing language and societal knowledge is supported andunderstood by most interviewees in all countries, some criticism exists. These criticalremarks mostly concern the content of the tests or study material, especially thequestions asked in the knowledge of society tests. These were often deemed not to berelevant, or even useless or primitive. Furthermore, in most countries various prob-lems and/or obstacles in the naturalisation procedure exist. For instance, in Denmarkit is considered a problem that the knowledge test is only held twice a year and inHungary the minimum level of required knowledge is unclear. In the Netherlands,the questions of the test are kept secret, which complicates the preparation for thetest.

ExemptionsIn all countries, except the United Kingdom, immigrants who have followed educa-tion in the country or who are attending school, are exempted from proving suffi-cient language skills and/or passing the knowledge of society test as a condition fornaturalisation. Persons who have reached a certain age are exempted from fulfillingthe language and knowledge of society requirements in Austria, France, Germany,Hungary, Latvia, the Netherlands and the United Kingdom. In all countries underconsideration, exemption from the requirements may be obtained if someone suffersfrom a mental or physical impediment.From interviews in Denmark, Latvia, the Netherlands and the UK it appears thatrespondents think that there are few possibilities for exemption and that exemptionis difficult to obtain.

PurposesIn the countries under consideration different reasons for introduction of the re-quirements and/or tests were expressed. The reasons mainly fell within three catego-ries. Firstly, in some countries the opinion on the place of naturalisation in the inte-gration process changed. In the Netherlands, for example, there is a shift from seeingnaturalisation as a step in the integration process to naturalisation as ‘the crown’ onthe integration process. Secondly, a more uniform interpretation of the language re-quirement for naturalisation was an important factor for formalizing naturalisationtests in Germany and the Netherlands. Finally, in many countries, the introductionand strengthening of the language requirements and the introduction of the condi-tion of societal knowledge have been justified by the argument that the requirementswould promote the integration of migrants..The situation in Latvia can be considered special because of the former Soviet oc-cupation and the need to ensure state continuity. In Latvia, naturalisation require-ments, especially the language requirement, were considered to be a basis for build-ing a civil society and for defining a shared system of values.

Effects on integrationFrom interviews conducted for this study, it appears that in most countries underconsideration it remains questionable whether naturalisation tests lead to better inte-gration in society. Most migrant respondents in Denmark, Germany, Latvia and theUK were of the opinion that the test did not contribute to their integration. Con-versely, other respondents pointed to the positive effects of naturalisation testing.Language teachers in Latvia, for example, said that examinations are necessary to in-tegrate in society because knowledge of language and history are crucial in daily life.In Denmark and the Netherlands most migrants experienced that attending acourse and the need to practice, increased their sense of independence. In most coun-tries, having passed the test is not recognised as a proof of certain skills on the labourmarket.

Groups facing difficulties fulfilling requirementsSome groups find it more difficult to meet the naturalisation requirements than oth-ers. Especially older people find it difficult to meet the naturalisation requirements inAustria, Latvia, the Netherlands and the UK. This is confirmed by data on the num-ber of naturalisations and pass rates broken down by age. In Hungary the high lan-guage skills necessary to pass the test on constitutional knowledge is the highest bar-rier for the ethnic minorities who do not speak Hungarian. In Latvia, younger peoplehave more problems with the knowledge examination. In the United Kingdom manypersons in humanitarian categories, some in family categories – especially if they arewomen – fare less well.

Chapter 5. Summary and ConclusionsTineke StrikA comparison of the developments in the nine Member States, clearly shows a grow-ing preference to impose integration conditions at the earliest possible stage and, inthis regard, to connect the right to residence with an integration requirement. In thischapter, we will analyse the political perception lying behind this development. Inthis analysis we included possible relations between the policy developments of thedifferent Member States. As this research has concentrated on the effects of the inte-gration measures and conditions, we will describe in this chapter what the relevantconclusions regarding the different tests have in common. Finally, we will formulatesome recommendations for best practice as well as for future research.

In most Member States the first type of integration requirement has been applied in anaturalisation procedure. Until the beginning of this century, this condition was li-mited to knowledge of the language. In most countries the fulfilment of this re-quirement was assumed (United Kingdom) or assessed in an informal interview witha civil servant (Netherlands, some German states, Denmark, France). Since 2000 thetendency has grown in different Member States to apply uniform criteria and to as-sess the language requirement in the form of a formalised test, often combined withan assessment on societal knowledge. Currently, seven of the nine Member States as-sess the required knowledge by a test.At the same time politicians have called for the need to prove knowledge of andrespect for the national values and norms. In France, the interview was not replacedby a test but the French values and norms have become increasingly important in theassessment. The new requirement for future French citizens to show adherence tothem, indicates that the emphasis on assimilation in French integration policy hasgrown. Considering the growing attention paid to national values and norms in theintegration policies of other Member States, this could be a common trend.Before 2000, naturalisation had been seen as a step in the integration process,which supported migrants to become more integrated. For this reason, a large num-ber of politicians were convinced at that time that granting citizenship to migrantsliving permanently in their country, was in the national interest. At the beginning ofthe twentieth century this conviction made way for the shift towards the idea thatmigrants should deserve citizenship, as it marked the final stage of the integrationprocess. This changing approach to migrants not only affected the requirements forcitizenship, but also those for residence rights.Since 2000 national governments have started to require a certain language levelfrom applicants for permanent residence. Germany has a longer history in this re-spect: a simple knowledge of the German language has been required for a settle-110

ment permit since 1990. Although language courses were already available in mostcountries, from the middle to the end of the 1990s the offer of courses became morestructured and extensive. The overall view was that authorities had to support mi-grants in learning the language. There was a consensus for the idea that not onlywould language skills promote integration, but also that access to work, the right tofamily reunification, permanent residence and citizenship would foster integration.Relatively soon after the introduction of a more structured education regime, newlyarrived migrants became obliged to participate in education programmes. In mostcountries this requirement to make an effort to integrate was the first step towards anobligatory integration policy. Not attending a course initially led at most to a reduc-tion in social welfare benefits. In Denmark the government started to offer an intro-ductory programme for newly arrived migrants in 1999. If they did not attend thecourse, their application for permanent residence could be refused. The efforts of amigrant to learn the language, hence became connected to his/ her residence rights.The next step towards a connection between integration and residence rights, hasbeen the development of an integration test for the acquisition of a permanent resi-dence right (currently, in seven of the nine Member States a test is taken for this pur-pose). The obligation to attempt to learn the language or learn about society, wasnow replaced by the obligation to achieve a certain result.After acceptance of the idea that a migrant must first show a certain level of inte-gration before having the right to permanent residence, politicians thought it wouldbe strange to require less from a migrant who applied for a stronger legal position, ci-tizenship. Although in most Member States the integration requirements were firstintroduced in the naturalisation procedure, the introduction of integration require-ments for permanent residence rights seems to have strengthened the integration re-quirement for citizenship in Denmark, Austria, Germany and the Netherlands.Furthermore, the introduction of integration requirements for residence rightsseems to have promoted the need for the introduction of integration requirements atan earlier stage, namely in the admission procedure. By this last step, the integrationpolicy has become connected with immigration policy, whereby the migration ruleshave been adapted for integration purposes. In 2011, knowledge of language (and ofsociety) will be required by five Member States as a condition for family reunifica-tion. According to the French rules, admission will be denied only temporarily (atmost two months), in order for the migrant to learn the language before arrival inFrance. The developments and the political debates clearly show that five MemberStates followed (or are going to follow) the example of the Netherlands, which firstintroduced this condition for admission.Also this new policy has emerged from the trend we previously described: before2000, in most Member States the dominant view was that the right to family reunifi-cation promoted integration of immigrants. Compared to a situation in which familymembers live separately in different countries, living as a family unit was assumed tobe more favourable to integration. For this reason the right to family reunificationhas been laid down in the EU rules for free movement of workers. This stance wasalso taken by the Member States during the negotiations on the Family Reunification

Directive.47Just before and after the adoption of the directive, political support grewfor the idea that family reunification could pose a threat instead of a stimulus to inte-gration. Therefore, family members should first prove their prospects for integrationbefore being admitted.In a policy where a stronger legal status is no longer used as an instrument forthe improvement of integration, the responsibility for the integration lies more on theshoulders of the migrant. But between the Member States where migrants first haveto prove their integration before their residence rights are strengthened, there is alarge difference in the extent to which they still show their responsibility. An impor-tant indicator of the responsibility the Member State takes is the extent to which it of-fers education on language and societal knowledge to migrants. Germany (with itspolicy of ‘promoting and demanding’) and France (integration contract) explicitlyexpress the notion of shared responsibility, and Denmark also acts in this way by of-fering good quality courses for free. In the Netherlands, the notion of shared respon-sibility between the state and the individual migrant for his/her integration process,has lost political support. This attitude has been demonstrated by the reluctance tosafeguard language courses abroad and, more recently, by the announcement in theCoalition Agreement of the new right-wing government Rutte to withdraw its (fi-nancial and organisational) responsibility for the integration education in the Nether-lands.The political development described above, is applicable to Denmark, the UnitedKingdom, Germany, Austria, the Netherlands and to a lesser extent to France. Whatis striking in this regard, is that in some Member States (particularly Denmark andthe Netherlands) from the moment it was decided to strengthen the integration re-quirements (for naturalisation) or introduce them (for permanent residence or admis-sion), the levels were quickly raised soon afterwards. Sometimes the requirementswere strengthened again one year after the previous strengthening. As these deci-sions could not have been based on evaluations of their effects, they apparentlyemerged from political preferences.

Three deviating Member StatesThe reports on Belgium, Latvia and Hungary show a different development, partlydue to a different historical background.The integration policy in Belgium deviates from the other western MemberStates, as no language or integration requirements have to be fulfilled in the naturali-sation procedure. Only the Flemish government applies an obligation to integrate tomigrants living permanently in Flanders, which is limited to the signing of an inte-gration contract and the attendance at courses. Failure to comply with this obligationcan lead to a fine, but does not affect the migrants’ residence rights. Though the Fle-47See Preamble of Directive 2003/86/EG, no. 4: ‘Family reunification is a necessary way of makingfamily life possible. It helps to create sociocultural stability facilitating the integration of thirdcountry nationals in the Member State, which also serves to promote economic and social cohe-sion, a fundamental Community objective stated in the Treaty.’

mish region tends to emphasise the importance of knowledge of language and socie-ty more and more, the overall Belgian policy still supports the notion that a success-ful integration of migrants primarily depends on social and inclusive policy, takinginto account the diversity of the Belgian population.The developments in Latvia and Hungary deviated from the western MemberStates. When Latvia regained its independence in 1991, after having been occupiedby the Soviet Republic for half a century, Latvian became the only official language.The requirements for Latvian citizenship were the topic of a political debate, whichwas closely related to the large group of Russian citizens living in Latvia. Due to alack of agreement, Latvian membership in the Council of Europe was postponed un-til the adoption of a new citizenship law. The High Commissioner on National Mi-norities of the Organisation for Security and Cooperation in Europe (OSCE) urgedthe government not to formulate language requirements for citizenship which wouldconstitute an obstacle to naturalisation. The level should not exceed conversationalknowledge, the authorities should be lenient in applying the requirements and elder-ly persons and the disabled should be exempted from the requirement. The law final-ly adopted nevertheless included the requirement to read and write Latvian and un-derstand official Latvian information. In 1998, pressure emerging from the process ofaccession to the European Union and low naturalisation rates led to the decision tosimplify the requirements for elderly persons and to the reduction in questions re-garding history and societal knowledge. At its implementation of the Long-termResidence Directive, the Latvian government decided to introduce a language re-quirement for obtaining this status. Although the level of this language requirementis lower (A2) than the level for Latvian citizenship (B1) and it does not include a re-quirement for societal knowledge, the number of applications for the LTR status isvery low compared to those for citizenship.The Latvian development shows that naturalisation in Latvia was, for historicalreasons, a sensitive issue. Certain politicians associated citizenship with loyalty, andwere hesitant to enable large numbers of Russians to become Latvian citizens. To thisend, the language requirement proved to be an effective instrument. One can con-clude somewhat ironically that pressure from the European Union led to relaxationof the language and societal knowledge criteria in a new Member State, whereas thedevelopment in the older Member States show an increasing level of requirementsfor citizenship. Considering the growing call for proof of adherence to Western val-ues and norms, loyalty has replaced integration as a national interest in the discus-sions on citizen- ship in the Western Member States.The introduction of the constitutional and societal knowledge requirements inthe Hungarian naturalisation procedure, besides language requirements, was justi-fied by a reference to other national practices in Europe. In this way, the governmentaimed to preserve and emphasise the constitutional, cultural and historical heritageof Hungary. From January 2011 onwards, ethnic Hungarians living outside Hungarywill be able to apply for Hungarian citizenship. Furthermore, applicants in Hungarywill be obliged to take an oath or pledge of allegiance before the local mayor of themunicipality of the migrant’s residence.

As the requirements or the strengthening of them have only been introduced rela-tively recently, it is too early to draw conclusions on the extent to which they havepromoted integration. Several rapporteurs but also many respondents emphasisethat successful integration, besides knowledge of language and society, depends onmany other factors as well. Furthermore, knowledge of the language of the MemberState is not always necessary to become integrated, especially if the migrant lives inan environment where another language, for instance English, is spoken.A Flemish evaluation showed that migrants who had accomplished their integra-tion course, more often acquired a job than the ones who did not finish the course.Danish research (Hansen 2009, see also paragraph 3.7), which found out that a moreefficient municipal integration policy resulted in more employment participation, in-dicates that the necessary support can take different forms.The above mentioned Flemish evaluation however also revealed that participa-tion in the course does not always result in a substantial improvement in the lan-guage skills or in social contacts with Flemish citizens. This outcome confirms the li-mitation of integration programmes, to which many respondents pointed. Accordingto Austrian respondents, a receptive society is crucial for their participation andhence, for their integration. These respondents missed a welcoming approach, andtherefore felt isolated.The interviews revealed that there is much support for learning about society be-fore departure to the Member State, but less for learning the language. Learning thelanguage however is perceived as the most valuable part of the courses after arrival.Respondents attending these courses, were more skeptical about the societal know-ledge they had to learn, for different reasons. Some found the content was not veryuseful, and others could not believe it would help them to integrate.The empirical part of the research shows a paradoxical conclusion with regard tothe different tests. On the one hand, the language level is perceived as insufficient toimprove the migrant’s labour market position. The level for admission is insufficientto act independently, the level for permanent residence is insufficient to be succesfulin the labour market. On the other hand, the level is perceived to be too high in orderto include all migrants who are willing to integrate.This latter perception has already been confirmed by the decline in the number ofapplications for family reunification (in Germany and the Netherlands), and will beexpected to result in an increasing number of denials of applications for a permanentresidence permit. The sharp decline in the number of permanent residence permitsgranted, which we currently see in Denmark, is expected in Austria, the Netherlandsand the United Kingdom as well. The current pass rates are influenced by the prac-tice that, until now, only the most motivated and able candidates have taken the test.In some countries, a side effect of the introduction of the test for permanent resi-dence rights is the introduction of language and integration courses – and vice versa.In general, migrants were positive about their participation in these courses, whichstrengthens their motivation to continue learning the language. Furthermore, partici-pating offers the opportunity for making social contacts, which prevents them fromisolation and promotes their integration as well. Furthermore, these contacts and114

How do the consequences of failing the test affect integration?The consequence for certain groups of being excluded from admission or a strongerlegal status, will not serve their integration, nor the integration as a general purpose.This is especially the case with regard to family reunification (where the marriage al-ready existed before the migrant in the Member State was granted a residence per-mit), Failing the test and therefore being denied access (to the Member State or astronger residence right) results in stress for both spouses, and prevents both of themfrom having a positive attitude to society. The delay in family reunification and theconcentration on trying to fulfil the criteria will also withhold the spouse living in theMember State from integration or participation. If children are involved, the delay inthe reunification will also affect their interests and their integration. The circums-tances of family members waiting for admission can be unsafe if they live in (former)war countries. Although spouses of recognised refugees are, according to the FamilyReunification Directive, exempted from the integration test, spouses who live in a(post) war situation like in Afghanistan or Sudan, face similar problems as familymembers of recognised refugees.If migrants retain their temporary residence permit because they cannot meet therequirements for the granting of a permanent residence permit, they will in generalhave fewer integration possibilities. Migrants with a temporary residence permit,depending on the Member State where they live, may be barred from full access tothe labour market or social benefits or from voting in local elections. The most im-portant consequence of not obtaining a permanent residence right is the continuationof insecurity regarding the residence right. If for instance an income requirement isno longer fulfilled, migrants may face withdrawal of their permit or expulsion. Thisconsequence is the most distressing for refugees or migrants who obtained a permiton humanitarian grounds. After all, they can be expelled to a country where they riskpersecution or arbitrary violence, which has perhaps caused trauma. Undoubtedly,this uncertainty will affect their prospects for integration. Though the integrationconditions are meant to promote integration, the opposite effect can be producedwith regard to migrants who are willing to integrate, but who lack the capacity topass the tests. Furthermore, in some Member States they may lose their temporaryresidence permit if they do not fulfil the integration requirements (for example Aus-tria, and in future the Netherlands).

Which groups are affected the most?As mentioned above, refugees (and migrants in a –post- war situation) and their fam-ily members can suffer relatively the most from being denied admission or a perma-nent residence right. Figures on pass rates by nationality and the interviews indicatethat they actually fail the tests relatively more often. This can be related to the harshcircumstances they live in abroad, or the trauma they suffer after arrival. This couldbe called an unintended effect, as most governments seemed to have other categoriesin mind when they introduced the integration tests. Most governments want to selectfor motivation and prospects for integration or to combat forced marriages. Thesepurposes are more related to family formation than family reunification, and to mi-grants who are free to choose where to live. This is not the case with refugees andtheir family members. These objections do by no means indicate that their integrationprospects are bad. A survey on the Danish local policy towards refugees and fami-lymembers indicates that the effectiveness of integration policy also depends on thequality of organization and the extent to which this policy is prioritized Hansen2009). The connection of the integration policy with their legal status is howeverproblematic and counterproductive, as a lower security of residence or less social orpolitical rights can affect the degree of integration.Other most affected groups mentioned by many respondents are elderly peopleand the low educated. These groups have, besides migrants with a refugee back-ground, the largest problems in meeting the integration criteria at all three stages.They lack experience with learning (a foreign language), or are not even familiar withthe latin alphabet. The duration of the course often appears to be too short for thelower educated and illiterates. Germany, the Netherlands and Flanders also obligemigrants who are already settled to meet integration requirements (passing the testor, like in Flanders, participating in a course). Elderly migrants, who have alreadylived in the Member States for a long time, also lack motivation, because, in theirview, the attention paid to their integration is too late. Another significant differenceseems to be that migrants originating from developing countries and non-Westerncountries have lower pass rates than migrants from developed and Western coun-tries. This could be related to a lower educational background, the lack of experiencewith computers or to the larger differences in culture and language between theMember State and their home country.As a matter of fact, all third-country nationals can be affected by these integra-tion requirements for the granting of a legal status, especially because an individualconsideration of the circumstances and interests is lacking. This development outdis-tances the position of third-country nationals from those of EU citizens, to whom in-tegration requirements are not applicable. Although the European Council an-nounced in 1999 that the rights of third-country nationals should become as compa-rable as possible to those of EU citizens, the developments have turned in the oppo-site direction. However, this so-called Tampere conclusion aimed to promote the in-tegration of third country-nationals.4848Conclusion no. 18. ‘The European Union must ensure fair treatment of third country nationalswho reside legally on the territory of its Member States. A more vigorous integration policy→

The research findings indicate that migrants benefit from language and integrationcourses, as it improves their language abilities as well as their social contacts, theirindependence and self-confidence. In general, migrants are highly motivated to par-ticipate in language education. They do not seem to be motivated by the obligation toattend a course, but because they like to learn the language and get in touch withother participants. For some specific groups, the obligation to attend the course canhave a positive effect on their motivation or actual participation. This particularlyconcerns women who are not supported by their husbands to integrate and migrantswho face difficulties combining the courses with their daily activities, such as a full-time job or raising and taking care of young children.However, especially if attendance to integration education is obligatory, moreflexibility is needed in order to take the individual circumstances of migrants into ac-count. This is with regard to the organisation of courses, for instance, more eveningcourses for migrants who are employed and preferably also more flexibility by theiremployers to grant them time for education. With regard to young parents, child careat the education centre facilitates their attendance. Also more flexibility towards theexemptions from the requirements is advisable. This is in relation to migrants whocan clearly demonstrate sufficient knowledge of language or society, as well as mi-grants who are not able to meet the requirements. Finally, flexibility towards the lev-el of the education offered is necessary. As the knowledge level and the ability tolearn differ greatly, sufficient differentiation in the education offer is crucial for themotivation and progress of the participants. Special attention should be paid to pos-sibilities to better attune integration education to regular education or the labourmarket. Therefore language education, combined with the development of specialcourses for migrants who lag behind but also for those who want to make moreprogress, would promote an effective integration policy. This is the case for languageeducation abroad as well as education in the country. The research clearly showsthat especially young migrants who have arrived recently in the Member States, ben-should aim at granting them rights and obligations comparable to those of EU citizens. It shouldalso enhance non-discrimination in economic, social and cultural life and develop measuresagainst racism and xenophobia’, Presidency Conclusions, Tampere European Council 15 and 16October 1999, SN 200/1/99.

efit from language and integration education. It could be reconsidered whether it isuseful to oblige older migrants (mostly the ones who have resided in the MemberStates for a long time) to attend the courses and pass a test. It would probably bemore efficient and adequate to offer elderly migrants education and tailor madesupport for their integration, without obliging them to attend courses.This research, however, did not find any reason to promote the connection of theintegration requirements with the granting of a certain legal status (admission, per-manent residence or citizenship). This connection is not necessary to motivate mi-grants, and it inevitably leads to the exclusion of certain groups from a secure legalstatus. The effects of this exclusion will hamper their integration, rather than promot-ing it. With regard to the integration requirement for admission, the exclusion alsoresults in affecting family life and violating the right to family reunification, as laiddown in the Directive on the right to Family Reunification. In most Member States,this connection between the level of integration and the granting of certain residencerights has been made relatively soon after the introduction of a language educationinfrastructure. Although the added value of this connection has not been proven,most respondents emphasised its positive effect on integration of the migrantsshould not be overestimated. As the negative side effects of this connection are al-ready clear, it is recommended that this policy is reconsidered.Furthermore, the respondents demonstrated a broad consensus about the limitedeffect of language and integration policy on the actual integration of migrants. Otherfactors, such as a receptive society, an effective combat of discrimination and equalopportunities on the labour market, are just as or even more crucial. To be effective,integration policies should pay attention to these elements.

More research necessaryIn order to monitor closely the effectiveness but also the side effects of the introduc-tion of the tests, it is recommended that research is conducted on the situation of mi-grants who are not able to meet the integration requirements, and are therefore pre-vented from admission, permanent residence or citizenship. Until now, little isknown about their circumstances and choices. As more Member States intend to in-troduce the language requirement for admission, and the effects of the integrationrequirement for the granting of permanent residence rights will become clear in thecoming years, this would be the proper moment to initiate such empirical research.The European Union has a long tradition in integration policy without the use oftests for the granting of a legal status. It could be worthwhile to assess the factorswhich led to a successful integration of EU citizens who made use of their freedom ofmovement. Looking at this experience, but also to best practices regarding the inte-gration of third-country nationals, may be more effective than continuing to cumu-late and strengthen the integration conditions.

BibliographyBauböck, R., E. Ersbøll, K. Groenendijk & H. Waldrauch (2006),Acquisition and Loss ofNationality, Volume I: Comparative Analyses: Policies and Trends in 15 EuropeanCountries,Amsterdam: University Press.Böcker, A., K. Groenendijk & B. de Hart (2005), ‘De Toegang tot het Nederlander-schap. Effecten van twintig jaar beleidswijzigingen’,Nederlands Juristenblad80(3)p. 157-164.Bonjour, S. (2010), ‘Between Integration Provision and Selection Mechanism. PartyPolitics, Judicial Constraints, and the Making of French and Dutch Policies ofCivic Integration Abroad’,European Journal of Migration and Law,no. 3, p. 299-18.Bundesregierung Deutschland (2010),Bericht über die Evaluierung des Nachweises ein-facher Deutschkentnisse beim Ehegattennachzug nach dem Aufenthaltsgesetz: Sprach-lern- und Sprachtestangebote, Visumverfahren,Berlin: Auswärtiges Amt, Beauftragteder Bundesregierung für Migration, Flüchtlinge und IntegrationBundesregierungÖsterreich,Nationaler Aktionsplan für Integration, Massnahmen,19 January 2010,www.integration.at.Carrera, S. (2009a),In Search of the Perfect Citizen? The Intersection between Integration,Immigration and Nationality in the EU,Leiden/Boston: Martinus Nijhoff.Carrerra, S. (2009b), ‘Nationality, Immigration and ‚the Republican Integration‛ inFrance: Normativisation, Expansionism and Externalisation’, in E. Guild, K.Groenendijk & S. Carrera,Illiberal Liberal States: Immigration, citizenship andIntegration in the EU,Burlington, VT: Ashgate, p. 315-336.Carrera, S. & A. Wiesbrock (2009),Civic Integration of Third-Country Nationals. Nation-alism versus Europeanisation in the Common EU Immigration Policy,www.ceps.eu.CERD (2010),Report on the Netherlands,Committee on the Elimination of Racial Dis-crimination, Seventy-sixth session 15 February to 12 March 2010,CERD/C/NLD/CO/1718.Commission (2008),Report from the Commission to the European Parliament and theCouncil on the Application of Directive 2003/86/EC on the Right to Family Reunifica-tion,COM (2008) 610, 8 October 2008.Cournil, C. & Y. Depigny (2008), ‘Contractualisation et externalisation de la politiquemigratoire : analyse et critique de la loi Hortefeux’,Revue de droit public,no. 4, p.1045-1079.Cuyper , P. de (2010),Deel 2. Het inburgeringsbeleid in Vlaanderen: de efficiëntie en effec-tiviteit geëvalueerd.Leuven: HIVA.Dagevos, J. & M. Gijsberts (2009),‘Ontwikkelingen in integratie op hoofdlijnen’,in J. Da-gevos & M. Gijsberts,Jaarrapport Integratie 2009,The Hague: SCP.ECRI (2008),Third periodical report on the Netherlands,Strasbourg, European Commis-sion against Racism and Intolerance.Ersbøll, E. (2010), ‘On trial in Denmark? ‘, in R. van Oers, E. Ersbøll and D. Kosta-kopoulou,A Re-definition of Belonging? Language and Integration Tests in Europe,Leiden: Martinus Nijhoff Publishers, p. 107-148.

Groenendijk, K. (2010),Predeparture Integration Strategies in the European Union: Inte-gration or Immigration Policy?,Washington DC: Migration Policy Institute.Groenendijk, K., E. Guild & H. Dogan (1998),Security of Residence of Long-Term Mi-grants. A Comparative Study of Law and Practice in European Countries,Strasbourg:Council of Europe.Hammarberg (2009),Report by the Commissioner for Human Rights Thomas Hammarberg,on his visit to the Netherlands,21-25 September 2008, 11 March 2009.Hansen, L.M. (2009),Effectiveness measurement of Danish municipalities integration poli-cies from 1999 to 2007,Copenhagen: The Ministry for Immigration, Refugee andIntegration Affairs.Hart, B. de & R. van Oers (2006), ‘European Trends in Nationality Law’, in R.Bauböck, E. Ersbøll, K. Groenendijk & H. Waldrauch (eds.),Acquisition and Loss ofNationality; Policies and Trends in 15 European Member States. Volume I: ComparativeAnalyses,317-357. Amsterdam: Amsterdam University Press. Also available viahttp://www.law.ed.ac.uk/citmodes/files/NATACCh7.pdf.Integrationsbeauftragte (2010),8. Bericht der Beauftragten der Bundesregierung für Mi-gration, Flüchtlinge und Integration über die Lage der Ausländerinnen und Ausländerin Deutschland,Berlin: Integrationsbeauftragte.Joppke, C. (2007),Do obligatory civic integration courses for immigrants in WesternEurope further integration? Focus Migration,Policy Brief no. 8. www.focus-migration.de/uploads/tx_wilpubdb/PB08_IntegrationCourses_02.pdf.Klekowski von Koppenfels (2010), ‘Citizenship Tests could signal that EuropeanStates perceive themselves as Immigration Countries’, in R. Bauböck & C. Joppke(eds.),How Liberal are Citizenship Tests,11-14. EUI Working paper accessible viahttp://eudo-citizenship.eu/docs/RSCAS_2010_41.pdf.Michalowski, I. (2007),Integration als Staatsprogramm: Deutschland, Frankreich und dieNiederlande im Vergleich.Münster: LIT Verlag.Michalowski, I. (2010) ‘Integration tests in Germany. A communitarian approach?’, inR. Van Oers, E. Ersbøll and D. Kostakopoulou,A Re-definition of Belonging? Lan-guage and Integration Tests in Europe,Leiden: Brill Publishers, p. 185-210.Ministerie van Justitie (2010),Rapportage Vreemdelingenketen. Periode januari-juni 2010.Den Haag: Ministerie van Justitie.Oers, R. van, E. Ersbøll & D. Kostakopoulou (2010),A Re-definition of Belonging? Lan-guage and Integration Tests in Europe,Leiden: Martinus Nijhoff Publishers.Oers, R. van (2006),De Naturalisatietoets Geslaagd? Een Onderzoek naar de Totstandko-ming en Effecten van de Naturalisatietoets,Nijmegen: Wolf Legal Publishers.Odé, A. (2009), De Wet Inburgering Buitenland: zelfselectie belangrijker dan selectie,Migrantenrecht2009, no. 7, p. 288-292.Pascouau, Y. (2010) ‘Integration Measures in France: An Evolving Process betweenIntegration and Migration Issues’, in: R. van Oers, E. Ersbøll & D. Kostakopou-lou,A Re-definiton of Belonging? Language and Interation Tests in Euope,Leiden:Brill, 153-83.Pauwels, F. & M. Lamberts (2010),Deel 3. De impact van inburgering,Leuven: HIVA.Pop, V. (2009), ‘Austria eyes language tests for migrants’,EU Observer,16 January2009, http://euobserver.com/9/29165.

Ratia, E. & Walter, A. (2009),International Exploration on Forced Marriages. A study onlegal initiatives, policies and public discussions in Belgium, France, Germany, the UnitedKingdom and Switzerland,Nijmegen: Wolf Legal Publishers.Regioplan (2009),De Wet Inburgering Buitenland, een onderzoek naar de werking, de resul-taten en de eerste effecten,publication no. 1754, www.regioplan.nl or TK 2008-2009,32 005, no. 1.Rother, N. (2009),Das Integrationspanel. Entwicklung von alltagsrelevanten Sprachfertig-keiten und Sprachkompetenzen der Integrationskursteilnehmer während des Kurses.www.bamf.de.Ryan, B. (2009), ‘The immigration Agenda in British Migration Law’, in E. Guild, K.Groenendijk & S. Carrera,Illiberal States: Immigration, Citizenship and Integration inthe EU,Burlington, VT: Ashgate, p. 277-298.Seveker (2008),Sprachkurse im Kontext von Integration und Aussiedlung. Am Beispiel derAltairegion/Westsibirien,Saarbrücken: VDM Verlag Dr. Müller.Significant (2010),Evaluatierapport Inburgering in Nederland.www.rijksoverheid.nl.Spijkerboer, T. (2007)Zeker weten: Inburgering en de fundamenten van het Nederlandsepolitieke bestel,Den Haag: Sdu.Walter A. (2008),Reverse Discrimination and Family Reunification,Nijmegen: Wolf Le-gal Publishers.Willems, W. (2009), De politiek aan de knoppen van de machine: spraaktechnologiein het inburgeringsbeleid’, in Huub Dijstelbloem & Albert Meijer (eds),De migra-tiemachine: de rol van technologie in het migrantenbeleid,Amsterdam: Van Gennep,p. 123-156.Winden, J. van der (2006),Wet inburgering in het buitenland,Den Haag: Sdu.

Chapter 1. Research Purpose and Methodology

Chapter 2. Integration Tests Abroad

Chapter 3. Integration Tests in the Country

Chapter 4. Integration Tests in the Naturalisation Procedure

5

11

45

77

Chapter 5. Summary and Conclusions

Bibliography

National Rapporteurs

109

119

123

1.1

Purpose of the Study

1.2

Methodology

1.3

Interviews

2.1

Organisation and Costs

2.2

Level and Content

2.3

Target Groups, Exemptions

2.4

Purpose of the Test and Debate

2.5

Consequences of Not Passing the Test

2.6

Effects of the Test: Statistical Data

2.7

Effects of the Tests: Interview Results

2.8

Summary and Conclusions

3.1

Description of the Requirements

3.2

Preparation for the Examinations

3.3

Costs

3.4

Target Groups and Exemptions

3.5

Consequences of not passing the test

3.6

Purposes and Debates

3.7

Effects of the Requirements: Statistical Data

3.8

Effects of the Requirements: Interview Results

3.9

Summary and Conclusions

4.1

Description of the tests

4.2

Preparation for the Tests and Costs

4.3

Exemptions

4.4

Purpose of the Tests

4.5 Consequences of not passing the test

4.6

Effects of the Tests: Statistical Data

4.7

Effects of the Tests: Interview Results

4.8

Summary and Conclusions

5.1

From an Effort Obligation to a Result Obligation: An Overview

5.2

Do Integration and Naturalisation Tests Improve Integration?

5.3

Recommendations

Co-benificiaries of the Intec project as funded by the Integration Fund:

Other national rapporteurs