10 European integration, consensus politics and family migration policy in Belgium and the Netherlands

Transcription

1 European integration, consensus politics and family migration policy in Belgium and the Netherlands Introduction Maarten Vink, Saskia Bonjour and Ilke Adam This chapter aims to investigate the relationship between European integration and consensus politics in the domain of family migration policy in Belgium and the Netherlands. We are interested both in how domestic traditions of consensus politics have affected policy- making on family migration at the European level, and in how the establishment of EU legislative competence has influenced the domestic dynamics of policy- making on family migration. As asylum flows towards Europe have declined since the end of the s, family migration has become the major channel for legal migration to many European countries. Family migrants are admitted on moral grounds, rather than selected on economic grounds. Family migration has therefore increasingly been cast as a problematic immigration flow, which should be controlled. Today, family migration is among the most salient topics on the migration agenda of the European Union and its member states (Boswell and Geddes 0; Bonjour 0; Bonjour and Vink 01). To what extent have these family migration policy dynamics been shaped by consensus politics? In line with the editorial introduction of this volume, we define consensus politics as formal and informal decision- making arrangements aimed at making decisions in a cooperative style involving as many political actors at the elite level as possible and avoiding decisions by simple majorities. A crucially important feature of this definition is that consensus politics is only a procedural arrangement, and does not imply substantive consensus. Rather, consensus mechanisms are aimed at accommodating fundamental political disagreement; where there is strong substantive consensus, consensus politics is superfluous. As a starting point for our analysis we make the observation that, while the Belgian and Dutch cases have much in common, the contemporary political dynamics in the field of family migration policies is strikingly different between them. In Belgium, there is no clear voting pattern with regard to family migration laws: some laws have been accepted by large coalitions, others follow the seat distribution between government and opposition. In the Netherlands by contrast, despite increasing politicization, substantive consensus has remained the 0 European ch.indd 1 0//1 1:1:00

2 1 M. Vink et al. norm. These different political dynamics in the Netherlands and Belgium allow us to analyze both the effect of consensus politics on European integration, and the effect of European integration on consensus politics. Before testing the propositions outlined by the editors of this volume for our two cases, we provide a brief background of post- war, policy- making dynamics in the domain of family migration policy. Post- war family migration policies The history of family migration policies in post- war Europe began in the 10s, with the migration of organized labor to Western Europe. During the guest worker experience of the 10s and 10s, states varied as to the extent that family migration was encouraged. In general, however, family migration policy was largely an administrative affair and not the subject of major political debate. The rights of migrant workers to reunite with their families left behind in the country of origin were determined by administrative guidelines, rather than by formal laws. In Belgium, family reunion was encouraged as a way of increasing the attractiveness of Belgium as a recruiting country in competition with the nearby French and German coal industry, and countering its declining birth rates, especially in Wallonia (Panciera and Ducoli 1; Rea 000). The brochure Living and Working in Belgium, for example, addressed potential North African migrants and praised Belgium s family reunification policy. Following a decree of 1, the government promised to reimburse half of the travel costs of spouses and children reuniting with labor migrants in Belgium. In contrast, in the Netherlands, family migration policy was relatively strict during the years of large- scale labor recruitment. The Ministry of Justice was concerned about over- population and housing shortages, and was convinced that once their families had come to the Netherlands, foreign workers would not return to their home countries. Between 1 and 10, family reunification was forbidden. This policy, however, met with criticism from the Ministry of Social Affairs: the foreign workers so urgently needed by Dutch business should be offered the possibility to bring their families over, lest they choose to go to another country instead. This plea found support in Parliament with the Conservative Liberal party (VVD) and the confessional parties, who thought it morally wrong to keep fathers and husbands separated from their families. In the course of the 10s, policies were gradually relaxed. By 10, labor migrants were allowed to bring their families over after one year of residence, provided they had adequate housing and a labor contract for at least another year. In the second half of the 10s and early s, Dutch conditions for family migration were relaxed further. Non- marital relationships, both heterosexual and homosexual, were admitted as grounds for entry and stay, and women were allowed to bring foreign family members over under the same conditions as men. In addition, settled migrants were gradually exempted from the income requirement in case of involuntary unemployment, while the residence rights of their European ch.indd 1 0//1 1:1:00

3 Migration policy 1 partners and children were substantially strengthened. This lowering of obstacles to family migration enjoyed broad parliamentary support: all political parties, except the small populist right Centrumpartij, considered family reunification to be virtually an absolute right. Since substantive consensus was the norm, consensus politics was redundant (Bonjour 00). As these rights were gradually formalized, and thus became less flexible, and as the need for migrant workers faded after the end of the post- war economic boom, the normative foundations of family migration policy became increasingly controversial. In Belgium, from, a new immigration law institutionalized the right to family reunion, but it was soon criticized by the Brussels francophone mayors from mainstream parties, who favored a more restrictive approach (Rea 000). The Gol Law of 1, named after the francophone Liberal Minister of Justice, resulted in a first significant restriction of family reunification policy. In the Netherlands, it was only in the early s that increasing public and political concern about migration and integration led to a restrictive turn in family migration policies. Based on the literature, we may expect non- majoritarian impulses in the field of migration policy in Belgium to be relatively weak. Consensual decisionmaking by means of grand coalitions is not the norm for all decisions; but it is for those decisions that threaten the survival of the consociation (Seiler 1; Deschouwer 00). In other words, consensus comes into play for decisions that directly affect the main cleavages, such as on linguistic, church state, or employers workers issues (Dewachter 001). Moreover, grand coalitions (which generally occur when the opposition supports majority proposals) can be the result of vote- or office- seeking strategies. As decisions on family migration policies are not expected to affect the stability of the country, a monistic government- versus-opposition mode of voting in Parliament seems likely to be the norm rather than a consensual oversized majority. In the Belgian monistic parliamentary culture, political agreement across cleavages is mostly crafted within the government, before laws are even presented to Parliament. This is especially the case with matters regarding the linguistic cleavage. Since the federal government represents the majorities of both linguistic groups, finding a consensus in the government requires a double majority in Parliament and its linguistic groups. This political consensus is crafted through governmental elite accommodation behind closed doors of intercabinet and inter- ministerial meetings. Consensual dynamics in migration policy- making have played a role due to different levels of politicization in the francophone and Flemish parts since the s (De Winter et al. 00). In Wallonia and Brussels, immigration is not strongly politicized. That is to say, while there is no strong anti- immigrant politicization, there is some pro- immigrant politicization, due to the importance of the immigrant electorate in Brussels (Adam 01a, 01b). In this context, we understand politicization as a process where those who have political power give increasing attention to an issue (Hermet et al. 1: 1). In Flanders, political parties operate in a party system where immigration is very much politicized. Thus, the immigration cleavage 0 European ch.indd 1 0//1 1:1:00

4 1 M. Vink et al. overlaps with the federal cleavage in Belgium. As a result, migration policies are always a compromise between very different views and are never radically restrictive or radically open. From that perspective, the consensual dynamic comes in through the back door, as it were. In the Netherlands, by contrast, consensus politics appears to be weak: decision- making in the field of family migration is strongly characterized by oversized majorities, not to overcome major conflict, but rather because of the absence of major conflict (Bonjour 0). Hypothesizing the effects of European integration In this section, we analyze the Europeanization of family migration politics in Belgium and the Netherlands, with specific attention to the two central questions of this volume: 1 To what extent has the domestic impact of European integration been mediated by consensus politics in Belgium and the Netherlands? To what extent has European integration affected consensus politics? Before discussing these questions, and some specific propositions, two preliminary remarks are required: one on the comparative design of this chapter and another on its substantive scope. First, as to the comparability of the Belgian and Dutch cases, most observers would probably agree that any differences between the two cases are the result, at least partially, of different institutional environments: whereas the unitary and hierarchical structure of Dutch migration policymaking would have facilitated policy reform in response to European integration, the decentralized federal structure of Belgium migration policy- making would have obstructed such reform. With regard to consensus politics, whereas we observe a strong substantive consensus in the Netherlands (between Left and Right) and thus weak consensus politics, in Belgium we observe the opposite: weak substantive consensus (between Flemish and francophone political parties) and strong consensus politics. As a result, our focused comparison of two cases in this chapter in some ways resembles a very similar systems design: two cases that have a lot in common (a general context of consensus democracy, strong politicization, Europeanization) but that differ in one crucial aspect (a politicized cleavage, which requires consensus politics to overcome conflict). Second, as to the role of Europe: for a long time it was out of the question that a matter as close to the heart of national sovereignty as entry and residence of third country nationals would be brought into the realm of the European supranational institutions. Only in the s, with the abolishment of internal border controls as part of the completion of the internal market, were the first careful steps towards European harmonization in this field taken. Initially, this was a strictly intergovernmental process. It was only after the Treaty of Amsterdam (1) that asylum and migration policies were gradually supranationalized that is, the European supranational institutions, the Commission, the Parliament and the Court, were granted a say. We focus in particular on the Family European ch.indd 1 0//1 1:1:00

5 Migration policy 1 Reunification Directive (EC 00/), which provides minimum norms with regard to the right of third country nationals to family reunification in EU member states. As a result of the directive, Dutch and Belgian family reunification policy has for the first time been subjected to Community law and to the scrutiny of the Commission and the Court. We now move to the elaboration of the main propositions that guide our analysis. We derive these from the editorial introduction of this volume and first formulate two hypotheses (HIa and HIb) about the way in which the impact of European integration has been shaped by characteristics of the domestic political system. In these propositions, consensus politics is viewed as an intervening variable. First, one could make the case that a centralized, hierarchical approach to policy- making facilitates the uploading of domestic preferences. This is probably one of the reasons why Europeanization has a centralization imperative in federal polities, as in the Belgian case (Beyers and Bursens 00b). As consensus politics implies a less hierarchical form of decision- making, one might expect that this has negatively affected the ability to upload domestic preferences to the European level. Hence, we investigate the hypothesis that consensus politics has negatively affected the ability to upload domestic preferences to the EU level (HIa). Second, in line with the idea of goodness of fit from the early Europeanization literature on administrative fit (Knill 001), one could expect that political systems with a strong culture of consultation and power- sharing do not easily adjust to synchronizing with a higher level political system, such as the EU. Hence we investigate the hypothesis that consensus politics has undermined the effective transposition and enforcement of EU legislation (HIb). We also formulate two hypotheses (HIIa and HIIb) about the way in which the domestic political system has been affected by European integration. In these propositions, consensus politics is viewed as a dependent variable. First, the very existence of EU policies may put pressure on national consensual politics. European integration offers monetary stability and military security. Particularly for smaller countries, this may weaken the idea of being vulnerable and thus undermine the incentive to practice consensus politics. This proposition would apply to situations where traditionally domestic consensus politics is needed to overcome major domestic conflicts. The imposition of European norms undermines those domestic incentives for special power- sharing mechanisms. Hence we investigate the hypothesis that European integration has weakened the incentive to practice consensus politics (HIIa). Second, the influx of immigrants has raised the salience of identity issues in the Low Countries, while European rules shape the context within which national policy- makers can operate. Given that European integration mainly undermines the ability to control domestic borders, and thus the inflow of migrants, one could hypothesize that this adds to the general tension between politicians and voters with cosmopolitan positions who look favorably on increasing geographical mobility and those who are more nationalist and view open borders as a threat (Kriesi et al. 00). Family migration policy would be a 0 European ch.indd 1 0//1 1:1:00

6 1 M. Vink et al. good example of a policy that has gone out of control. Hence, we investigate the hypothesis that Europeanization has induced political controversy over the control over national borders and has strengthened the political divide between nationalists and cosmopolitans (HIIb). Family migration policies since Whereas, previously, policy- makers paradoxically accepted unwanted immigration (Cornelius et al. 1; Joppke 1, 001; Cornelius and Tsuda 00), since the s accepting rights- based migration such as family migration or asylum has become much less self- evident and more contested. Particularly since the 000s, family migration policies have become a part of high politics. In this chapter, we investigate how consensus politics has shaped the impact of European integration in the Netherlands and Belgium and how European integration has impacted on consensus politics in these countries. Before evaluating the propositions described above, in this section we first discuss the context by outlining the developments since the s in the field of family migration policy in these countries. In the Netherlands, family migration policies have generally rested on very broad political majorities since the 10s, and this remains the case today (see Table.1). As a rule, reforms of family migration policies are supported by all political parties except those left of the Labor Party (PvdA): the Green Party (GL) and the Socialist Party (SP). The generally slightly smaller majorities witnessed in the 000s compared to the s, were primarily due to the increased electoral success of the Far Left (the Green Party and the Socialist Party together won seven seats in 1 and thirty- two in 00). There are two exceptions to the general trend that only the Far Left opposes reforms of family migration policies. The first is the Law on Sham Marriages of 1, which was opposed not only by the Green Party but also by the Liberal Democratic Party (D). The law allowed the Registrar ( Ambtenaar van de Burgerlijke Stand ) to refuse to contract a marriage if there were suspicions of fraud. The Liberal Democrats thought it inappropriate to involve the Registrar in the fight against the abuse of immigration law. The second exception was the increase of the level of income that sponsors of family immigration must earn, from 0 percent to 10 percent of the minimum wage in 00. This measure was opposed not only by the Green Party and the Socialist Party, but also by the Labor Party. Even though the Labor Party had previously implemented restrictive reforms of the income requirement (in 1 and 000) when they were in office, income requirements still tend to be a sensitive issue for the Labor Party, since restrictive reforms of the income requirement exclude those in the weakest socio- economic positions (Bonjour 00). There has also been continuity since the 10s, in that the broad political support for family migration policies is not generally the product of overcoming conflict through consensus politics, but of actual agreement between political parties. The early s were an exception to this rule. While the Labor Party European ch.indd 1 0//1 1:1:00

7 Migration policy 1 Table.1 Most important reforms of Dutch family migration policies, 0 Year Reform Government Opposed Seats in favor 1 Income requirement reintroduced (0% welfare level) CDA PvdA 1 Law on Prevention of Sham Marriages CDA PvdA 1 Family migration allowed only if Provisional Entry Permit (MVV) was obtained abroad 000 Income requirement raised to 0% welfare level 00 In case of family formation: income level raised to 10% minimum wage and minimum age raised to 1 years PvdA VVD D PvdA VVD D CDA VVD D 00 Law on Integration Abroad CDA VVD D 00 Level integration abroad raised from A1 to A1; reading & writing added to oral language test in integration abroad exam; requirement of integration and education for applicant CDA PvdA CU GL 1 seats (%) GL D GL SP Unie + GL SP PvdA GL SP GL SP 1 seats (%) 1 seats (%) 1 seats (%) 1 seats (1%) 1 seats (%) still stuck to the principle that was dominant in the s, that family reunification was an all but absolute right, the Christian Democrats had adopted the new point of view, that the social problems related to migrant integration justified imposing conditions on family migration. The reintroduction of the income requirement in 1, at the modest level of 0 percent of welfare, reflects the compromise between these two coalition partners, crafted in strenuous negotiations behind the closed doors of cabinet meetings (Bonjour 00). In the course of the s, however, political consensus was restored, as the Labor Party came to share the conviction that family migration was a conditional right. This shift in preferences towards the restrictive end of the spectrum occurred in all political parties. While the shift was gradual in the s, it took a radical turn in the 000s in response to the electoral successes of the Far Right GL SP seats (%) Sources: Neth. Doc. Sec. Chamber, 1 1, 0, 1; 1 000,, ; 1 000, 1, ; 00 00, 1, ; 00 0, 00, ; Neth. Proc. Sec. Chamber, 1 1, ; 1 1,, ; 1 000,,, p. ; 00 00,,. Note Where the reform consisted of legislative change, the two last columns in this table present the votes casted in the House of Representatives. Where it concerned a change in regulations, these columns represent the opinion of the political parties in the House as expressed in Committee or plenary meetings. 0 European ch.indd 1 0//1 1:1:00

8 00 M. Vink et al. (Pellikaan et al. 00; Bonjour 00). On the whole, politicization of migration policy in the Netherlands appears to have led to a substantive convergence, rather than a divergence, of standpoints on family migration. In Belgium, which has a consensus democracy like the Netherlands, decisionmaking on migration and family migration issues does not necessarily rest on grand coalition voting. Table. shows that a clear decision- making pattern with regard to family migration is less easily identified in Belgium than in the Netherlands. However, this does not mean that consensus politics has not been in play. Except for the 0 law, voted on during the long- term absence of a federal government, consensus politics took place in the form of compromises on the proposals crafted at the level of the federal government. Discussing Table. Most important reforms of Belgian family migration policies ( 0) Year Reform Government Opposed (*abstained) 1 Fixed term wherein administration must statute; min. age 1 years; control period on cohabitation 1 year (was: months) 1 Law on Prevention of Sham Marriages 00 Law following the directive: min. age spouses 1 years control period cohabitation 1 years family reunion unmarried partners and handicapped major children restriction spouse polygamous husband interdiction of waterfall family migration () is cancelled 00 Sham marriages as a criminal offence 0 Restriction family migration Belgians (same as rd country nationals); 10% income requirement; impossibility of family migration with ascendants; permanent permit after years (was: ). CVP PSC-PS SP Agalev-Ecolo CVP- PSC-SP-PS- VLD-MR MR VLD PS SP.A-Spirit MR-VLD PS-SP.A CD&V-NV-A Open VLD CD&V NV-A MR VLD* PRL VB Ecolo-Agalev VU-VB CD&V NV-A CdH Ecolo VB Ecolo VB PS* CdH* Groen Ecolo VB Seats in favor 1 seats (%) seats (%) seats (%) seats (%) seats (0%) Sources: Be. Proc. Chamber, 1 1,, pp. ; 1 1,, pp.,, and ; 00 00, 1, pp., ; 00 00, 1, pp., 0; 0 0,, pp. 11, European ch.indd 00 0//1 1:1:00

9 Migration policy 01 proposals before, or parallel to, parliamentary debates behind the closed doors of government meetings allows a double majority to be assured, in casu a simple majority in Parliament and a majority inside the linguistic groups. Since the mid- s, when the linguistic and migration cleavages started to overlap, decisionmaking behind the closed doors of government has allowed for the pacification of the linguistic migration conflict, resulting in family migration law that has been neither very restrictive, nor very liberal (until 0). We see this lack of transparency in decision- making as a typical manifestation of consensus politics, necessitated by a lack of substantive consensus. The support of the green parties for the 1 reform of family migration legislation must be interpreted in the following context. The year 1 was the year of the Sint Michiels Agreement, a major state reform, which was supported by the green parties in return for a tax on plastic packages. As with the family migration legislation, large majorities that are formed for the sake of a state reform can lead to other legislative changes supported by the same grand coalitions. Both the laws on the prevention (1) and the criminalization (00) of sham marriages have been supported by opposition parties, so this is seemingly a less conflictual policy issue than family reunion law as such. Whereas the 1 law was supported by both liberal parties, the 00 law was supported by the Flemish opposition parties CD&V and N- VA. The 00 amendment of the Foreigners Law implemented three European directives in the area of migration policy (on family migration, qualification for refugee status and victims of human trafficking) and included not only a reform of family migration policy, but also a major domestic overhaul of asylum procedures. The decision- making process leading to the family migration reform of 0 is an exceptional case of non- consensual politics. In the absence of a government since the federal election of June 1, 0, the Parliament accepted the law by an alternative majority of five Flemish parties and one francophone party. The constitution of this majority happened in the spotlight of the parliamentary arena, far away from the closed doors of government meetings. Especially in a domain as politicized as migration policy, this proactive and independent parliamentary activity, as well as the absence of a double majority in Parliament and in linguistic groups, is highly exceptional in Belgian political decisionmaking. How has consensus politics shaped the impact of European integration? The political dynamics of family migration in Belgium and the Netherlands lend empirical support to hypothesis HIa: that consensus politics has negatively affected the ability to upload domestic preferences to the EU level. In the Netherlands, consensus politics is weak in the field of family migration policy, not only because of the broad consensus among political parties, but also because the influence of interest organizations in the field of migration and integration is marginal (Scholten 00; Bonjour 00). In line with hypothesis H1a, 0 European ch.indd 01 0//1 1:1:00

10 0 M. Vink et al. this might have enabled the Netherlands to exert significant influence on policymaking, both at the EU level and in other member states. The negotiations on the Family Reunification Directive, which lasted from 1 until 00, were conducted on behalf of the Netherlands by the governments Kok II (Labor Party, Conservative Liberals and Liberal Democrats), Balkenende I (Christian Democrats, Conservative Liberals and the populist right wing LPF ) and Balkenende II (in which the Liberal Democrats replaced the LPF ). The Netherlands was one of the most influential member states in the negotiations: thirteen out of its seventeen proposals were adopted (Strik 0). The Dutch input into the negotiations was aimed at ensuring that the directive fit Dutch policies as closely as possible. With the support of Denmark, Germany and the UK, the Kok II government succeeded in eliminating the Commission s proposal that the directive forbid discrimination of a member state s own nationals over other EU citizens. Both the Kok and the Balkenende governments successfully pleaded for the directive to apply not only to married but also to non- married couples, including same- sex partnerships. The Balkenende governments added a number of points to the Dutch wish list, to ensure that the directive allowed the implementation of the measures foreseen in their coalition agreements, notably the increase of the minimum age for family reunification based on marriage from eighteen to twenty- one years of age and the introduction of a civic integration exam abroad. It took strong pressure from Austria, Germany and the Netherlands to introduce the clause allowing integration measures. The Dutch were initially the only delegation that wanted to raise the minimum age above eighteen years, but with last minute support from Belgium and Denmark, the Netherlands managed to have its way (Groenendijk and Minderhoud 00; Strik 0). Dutch governments have continued their lobbying at the EU level even after the adoption of the directive, pushing for other member states to copy the Dutch policy of having integration requirements for family migrants pressure that has led to such requirements being implemented in Austria, Denmark, France, Germany and the UK (Bonjour and Vink 01; Bonjour, forthcoming). Each of these measures has enjoyed broad domestic support, which has strengthened the position of the Dutch government in Brussels. 1 Migration being a sensitive issue in Belgian politics, partially overlapping with the linguistic cleavage because of a different politicization in both party systems, consensus politics is at play in the field of family migration. The consensual mechanism involved in reaching a Belgian position with regard to EU policy- making in general, and family migration in particular, might at least explain the perceived difficulty of Belgium in defending its national interests at the European level (Beyers and Bursens 00a: 0). Research on the Europeanization of the Belgian administration shows that Belgian civil servants dealing with EU policy largely agree with the idea that the complex Belgian state structure explains its difficulty in uploading its national interests (Baetens and Bursens 00). A Belgian diplomat adds that, since the Belgian position on migration issues already reflects a compromise (between sometimes very European ch.indd 0 0//1 1:1:00

11 Migration policy 0 divergent positions), it is less successful in uploading its preferences. The process of reaching a Belgian standpoint on a proposal for an EU directive is coordinated by the Directorate- General European Affairs (DEA) of the Belgian Federal Ministry of Foreign Affairs. The DEA invites all relevant administrations and ministerial cabinets concerned, federal and regional, as well as the cabinets of the four vice- premiers. The DEA is then charged with finding a consensus during these meetings. Besides the complex consensus decision mechanisms and the Belgian compromise positions that result from it, the pro- European attitude of Belgian civil servants has also been put forward to explain the low political input in EU policy- making (Beyers and Bursens 00a; Wessels 001). After all, by favoring a European compromise, bureaucrats from Belgium rely more strongly than others on the European Commission to represent their interests in the negotiations. During the negotiation process of the family reunion directive, Belgium was perceived as an active participant, but with a low capacity in uploading its national preferences (Strik 0). Its active participation might be explained by its EU presidency at a central time in the negotiation process of the directive (July December 001). During the 001 presidency, the Belgians teamed up with the Commission, the front- runners of a European harmonizing perspective. Belgium s active, yet presiding role made it difficult to consistently push for its national preferences. After the Belgian presidency, Belgium stopped its role as a front runner of European harmonization of family reunification policy (Strik 0: ). Our empirical findings only partially support hypothesis HIb, that consensus politics has undermined the transposition and the effective enforcement of EU legislation. Since the politics of family migration in the Netherlands has been characterized by weak consensus politics, the hypothesis would predict few if any transposition problems. What we find, however, is the opposite. The Balkenende II government happily put its signature under the Family Reunification Directive, convinced that it would not require a substantive change to Dutch policies. However in a 00 report, the Commission explicitly criticized the Dutch transposition of the Family Reunification Directive, inter alia for setting a high income requirement for young people, for failing to grant facilitated access to visas and for fighting marriages of convenience in a manner that reflected generalized suspicion (COM /00). Two years later, the Netherlands was condemned by the European Court of Justice. In its judgment on Chakroun v. Minister van Buitenlandse Zaken (0), the Court found that the Dutch income requirement of 10 percent of the minimum wage was in breach of the Directive. In March 0, a new case pertaining to the Family Reunification Directive was brought before the Court: once again it concerned the Netherlands (Imran v. Minister van Buitenlandse Zaken 0). The question put to the Court was whether the Dutch integration exam for family migrants was compatible with the Directive. The case was dropped without the Court giving a ruling, but not before the Commission had advised the Court to condemn Dutch policy (European Commission 0). 0 European ch.indd 0 0//1 1:1:00

12 0 M. Vink et al. In other words, the transposition of the Directive in the Netherlands has been highly problematic, in spite of the virtual absence of domestic opposition. In part, the transposition problems have been due to the fact that Dutch politicians did not foresee that the jurisprudence of the Court would turn a text that appeared to leave member states a great deal of leeway, into a precise and constraining legal instrument. In addition, however, the policy preferences of Dutch politicians across the political spectrum have shifted ever further towards restrictive family migration policies, far beyond the limits of what the Directive allows. Even the Center- left Balkenende IV government proposed measures which clearly would not have complied with the requirements of the Directive, such as introducing education requirements for applicants (Neth. Doc. Sec. Chamber, 00 0, 1, 1). Pressured by the populist right and unhampered by any need to craft political compromises, Dutch governments have introduced radically restrictive reforms of family migration policies, even at the risk of breaching EU law. Paradoxically, Belgium is a Europhile country (Beyers and Bursens 00a), but not the best pupil of the class when it comes to the transposition of EU policies (Bursens 00). The Family Reunification Directive was also transposed too late. The ministerial counselors who negotiated the law proposal attributed this late transposition to the fact that the law, regulating amongst others the new family migration policy, transposed three directives at once (EC 00/; EC 00/1; and EC 00/) and completely reformed the asylum procedure, which was particularly complex. The timing of the directives, all from 00 and implying a modification of the Foreigners Law, might have stimulated the government to negotiate the different amendments as a whole for practical reasons. Yet, this regrouping might also indicate that the consensus mechanism of the package deal was used. Agreeing on a number of related items may facilitate but prolong the decision- making process. According to the words of a negotiator for a minister from the Parti Socialiste, it seems that the whole reform of the law must indeed have been considered a package deal: [N]o, [changing the family reunion legislation was not part of our agenda] because we knew that, anyway, the other side [the Flemish Liberals, VLD] was asking for reform.... We were only on the demand side for subsidiary protection, which concerns the other directive. In this case, late transposition can be related to the use of a specific consensus mechanism, namely that of a package deal. Back door opposition, or the idea of deliberately infringing EU law to manifest dissatisfaction with the agreed European measures, seems quite unlikely in a Europhile country like Belgium. Yet, this is what happened in 0 when Belgium transposed the Family Reunification Directive. The 0 Belgian family migration law sets an income requirement of 10 percent of the minimum living benefit (below the level of the minimum wage) for sponsors of family reunion. This law was approved by Parliament with the knowledge that the European ch.indd 0 0//1 1:1:00

13 Migration policy 0 Netherlands had been condemned by the European Court of Justice precisely for having a strict income requirement (Chakroun v. Minister of Foreign Affairs 0). Although the Belgian income requirement is less stringent then the Dutch one, lawyers argue that this income provision is also not compliant with EU law (Kruispunt Migratie- Integratie 0). Searching the limits of what is admissible under EU law appears to be a deliberate strategy of the Belgian parliamentary majority, aimed at pushing the EU legislative institutions to adapt the directive in a more restrictive sense (Be. Proc. Sec. Chamber, 0 0,, p. ). The exceptional majority decision- making process of the 0 family migration law thus contributed to, instead of preventing, the back door opposition by deliberately seeking the limits of EU law. Contrary to hypothesis H1b, we think that playing with the limits of EU law would not have happened if there had been a traditional consensual decision- making process that had pre- decided matters behind closed doors with a double majority of both linguistic groups. These results suggest that consensus politics leads to moderate reforms, while majority politics leads to more radical reforms, more likely to breach the limits of EU law and, thus, resulting in transposition problems. The effect of European integration on consensus politics As to hypothesis HIIa, that European integration has weakened the incentive to practice consensus politics, we find that in the Netherlands the near unanimous support for restrictive reforms of family migration policies was already such that there was no need for consensus politics. The introduction of EU law did not change this fact. The situation in Belgium was more interesting. In Belgium, in 00, while Flemish parties were eager to restrict family migration, francophone parties, in particular the francophone Socialists, would not even have wanted to put restrictions on family migration on the agenda if it had not been for the European Directive. Europeanization thus bridged a sub- national ideological division by forcing an issue onto the political agenda. However, this surprisingly strengthened consensus politics, rather than undermining it. Once the issue was on the agenda and family reunion legislation had to be revised, both francophone and Flemish government parties played the consensus game. The agreement on the law proposal was made in inter- cabinet meetings and in the ministerial councils between the representatives of the different ministers. The main actors in these meetings were the cabinet of the Flemish Liberal Minister of Interior, Patrick Dewael in favor of a restrictive interpretation of the family migration policy and the cabinet of the francophone Socialist Minister of Justice, Laurette Onckelinx, who tried, with some success, to limit the damage. Not being able to avoid the legislative change because of the Directive stimulated the negotiators to play the compromise game. In response to the Flemish Christian Democrats in opposition, who critically observed that the new law did not reflect the minister s ideological position, the Minister of Interior recalled the necessity of consensus politics: 0 European ch.indd 0 0//1 1:1:00

14 0 M. Vink et al. If I can reach a consensus, Miss Lanjri, that allows the democratic majority to progress with law projects, which, in the future, will improve the procedure... then I go for it. The day that you can participate in government, and you say your personal conviction must prevail, and that you won t do anything if you cannot realize your personal opinion for a 0%, then you will not govern. We are [in the government] not only confronted with an ideological difference, but we are in a situation where two communities, north and south [of the language border] (switches to French) and I say this clearly think in a completely different way. You have to assess then, if at a certain moment [law] projects are sufficiently important to vote on them. (Be. Proc. Chamber, 00 00, 1, p. ) As to hypothesis HIIb, finally, that European integration has induced political controversy over the control over national borders and has intensified the political cleavage between nationalists and cosmopolitans, we find empirical support in the case of the Netherlands, and contradictory evidence in Belgium. In the Netherlands, Europeanization has become a double- edged sword since the conservative Rutte government, composed of Conservative Liberals and Christian Democrats with minority support from the populist right- wing Freedom Party, entered office in 0. The Support Agreement signed by these three parties listed no less than seven reforms of family migration policies, which could only be implemented if the Family Reunification Directive were modified. Striving to make these changes in the Directive was an integral part of the Support Agreement (Neth. Doc. Sec. Chamber, 0 0,, 1, p. ). The Freedom Party has been quite vocal in its rejection of any EU interference with Dutch immigration policies, since it believes that it is due to that club in Brussels that Europe is quickly turning into Eurabia (PVV 0: 1). The Conservative Liberals, who used to be warm supporters of a European migration and asylum policy, have come to criticize EU law and jurisprudence as an obstacle to national sovereignty, stating that we want a restrictive immigration policy. We want all elements in European regulations which obstruct such a policy eliminated before we even consider a huge new pack of rules from Brussels (Neth. Doc. Sec. Chamber, 00 0, 0,, p. 1). This contrasts with the Labor Party, who remain convinced that what we stand to gain (...) in a European framework is much bigger than if we would muddle on our own. They argue that the Netherlands should continue steering a European course, also if this means we can t have it our way completely, because that is European democracy (Neth. Doc. Sec. Chamber, 00 0,, 1, pp. 1, 1). To Prime Minister Rutte however, opting out of EU migration law is a serious option: he has refused to rule out that if we can obtain an exception, we shall do so (Neth. Proc. First Chamber, 0 0,,, p. 1). The populist Right consistently presents both immigration and Europeanization as external threats to the nation. Mainstream conservative parties, notably the Conservative Liberals which won the 0 elections, have partly taken over this agenda. Thus, the Europeanization of migration policy itself has become European ch.indd 0 0//1 1:1:00

15 Migration policy 0 subject of political controversy, feeding into the cleavage between nationalists and cosmopolitans. In Belgium, overall support for European integration is stronger and there is a less clear relation between the nationalist globalist cleavage and the issue of migration. In Belgium only the extreme right- wing Vlaams Belang and the smaller Front National have an outspoken nationalist and Eurosceptical position (Szczerbiak and Taggart 00). The Flemish N- VA (as well as, in the past, the Volksunie) has a more constructive attitude towards European integration because it views the European Union as a stable framework within which regional independence for Flanders could be achieved (Ferreira Antunes 0). In other words, the strong contrast between nationalists and cosmopolitans in Belgium does not seem to be influenced by the Europeanization of migration policies. Reflections on European integration and consensus politics This chapter aimed to investigate the Europeanization of family migration politics in Belgium and the Netherlands, with specific attention to two general questions: How have domestic traditions of consensus politics affected policy- making at the European level? And, how has the establishment of EU legislative competence influenced the domestic dynamics of policy- making in countries with a tradition of consensus politics? Following the editorial introduction of this volume, we formulated four hypotheses, which we subsequently used to guide our empirical analysis of the two cases. When we look at the overall evidence, it quickly becomes clear that the picture is ambiguous, at best. Let s first look at the two bottom- up hypotheses. With regard to the hypothesis that consensus politics has negatively affected the ability to upload domestic preferences to the EU level (HIa), our findings are broadly supportive. The fact that the Netherlands was more successful than Belgium in shaping the Family Reunification Directive to match their policy practices is explained in part by the fact that the Dutch input into the negotiations was formed by centralized decision- making and based on broad political support, while the Belgian position reflected a strenuously crafted compromise. With regard to the hypothesis that consensus politics has undermined the effective transposition and enforcement of EU legislation (HIb), the evidence from our chapter points mainly against this proposition. Even though consensus politics in the field is quite weak in the Netherlands, the Chakroun case clearly indicates transposition problems. In Belgium the Directive was transposed too late (00) and the necessity of a package deal as a mechanism to reach a consensus seems to have played a role in this late transposition. However, in both our cases, we observe that the absence of consensus politics may result in radical reform, with politicians deliberately seeking out the limits of European law or even crossing those limits. In Belgium in 0, this back door opposition to the Directive was a reflection of an (exceptional) case of majoritarian decisionmaking. The limited validity of hypothesis HIb, which follows an administrative 0 European ch.indd 0 0//1 1:1:00

16 0 M. Vink et al. goodness of fit logic, in our empirical cases, is in line with increasing criticism of the goodness of fit hypothesis in the state of the art: what matters are domestic coalitions and institutions, rather than the fit or misfit between European and national policies (Mastenbroek and Kaeding 00). With regard to the hypothesis that European integration has weakened the incentive to practice consensus politics (HIIa), we find evidence against this proposition in the Belgian case. In Belgium the Family Reunification Directive forced family migration policy reform onto the political agenda and thus stimulated politicians to play the compromise game as they were unable to avoid legislative reform because of the European Directive. In the Netherlands, however, where there is no tradition of consensual politics in the field of family migration policy to begin with, Europeanization had no such effect. Europeanization has neither weakened nor strengthened consensus politics. With regard to the hypothesis that Europeanization has induced political controversy over the control of national borders and has intensified the political cleavage between nationalists and cosmopolitans (HIIb), we find empirical support for this proposition in the Dutch case, but not in the Belgian case. In the Netherlands, the Europeanization of migration policy itself has become subject to political controversy, feeding into the cleavage between nationalists and cosmopolitans. In Belgium on the other hand, there is little controversy over Europe and the Europeanization of Belgium. The Vlaams Belang is Eurosceptical, but receives little support for that position. The NV- A plays the European game to strengthen its sub- state nationalist project and favors the harmonization of European migration policies. Our research results are most clear- cut with regard to the effect of consensus politics on Europeanization. In both the Dutch and Belgian cases on family migration policy, the domestic impact of European integration seems to have been mediated by consensus politics and we observed a negative effect of consensus politics on the ability to upload national preferences. As for the downloading capacity of EU law, we observed that the absence of consensus decision- making can be a potential hindrance for the effective transposition of EU law, as it stimulates opposition to EU policies through the (transposition) back door. Thus, while consensus politics appears to make it more difficult for domestic actors to play the European game at the EU level, domestically it might facilitate accurate transposition. Our research results are more ambiguous with regard to the effect of Europeanization on consensus politics, although they suggest that Europeanization is more likely to reinforce consensus mechanisms rather than weaken them. Whereas in Belgium the incentive to practice consensus politics was strengthened due to Europeanization, in the Netherlands Europeanization has increased the politicization of migration policy, and thereby increased the need for consensus politics to bridge the emerging nationalist cosmopolitan cleavage European ch.indd 0 0//1 1:1:01

17 Notes Migration policy 0 1 Strik (0) points at additional factors, such as expertise and the language ability of Dutch negotiators. Interview with a Belgian diplomat, December 1, 01. Interview of March, 0. Own translation. Interviews with francophone and Flemish negotiators of the government law proposal to implement the directive, February and March, 0. Interviews with francophone and Flemish negotiators of the government law proposal to implement the directive, February, and March, 0. Primary sources Belgian Federal Parliament. Chamber of Representatives. Proceedings Chamber, 1 1,. Belgian Federal Parliament. Chamber of Representatives. Proceedings Chamber, 1 1,. Belgian Federal Parliament. Chamber of Representatives. Proceedings Chamber, 00 00, 1 1. Belgian Federal Parliament. Chamber of Representatives. Proceedings Chamber, 00 00, 1. Belgian Federal Parliament. Chamber of Representatives. Proceedings Chamber, 0 0,. Belgian Federal Parliament. Wet van 1 September 00 tot Wijziging van de Wet van 1 December Betreffende de Toegang tot het Verblijf, de Vestiging en de Verwijdering van Vreemdelingen. Belgisch Staatsblad, 00,. Belgian Federal Parliament. Wet van Juli 0 tot Wijziging van de Wet van 1 December Betreffende de Toegang tot het Verblijf, de Vestiging en de Verwijdering van Vreemdelingen Wat Betreft de Voorwaarden tot Gezinshereniging. Belgisch Staatsblad, 0, 1. Bibi Mohammed Imran v. Minister van Buitenlandse Zaken, C- 1/ PPU (0). Chakroun v. Minister van Buitenlandse Zaken, C- /0, ECR I- 1 (0). Commission of the European Communities. Report from the Commission to the European Parliament and the Council on the application of Directive 00//EC on the Right to Family Reunification. Register of Commission Documents, COM (00): /f. Council of the European Union. Council directive 00//EC on the Right to Family Reunification. Official Journal of the European Union, L- 1 (00): 1. Council of the European Union. Council directive 00/1/EC on the Residence Permit Issued to Third Country Nationals Who Are Victims of Trafficking in Human Beings or Who Have Been the Subject of an Action to Facilitate Illegal Migration, Who Cooperate with the Competent Authorities. Official Journal of the European Union, L- 1 (00): 1. Council of the European Union. Council directive 00//EC on the Minimum Standards for the Qualification and the Status of Third Country Nationals and Stateless Persons as Refugees or as Persons Who Otherwise Need International Protection and the Content of the Protection Granted. Official Journal of the European Union, L- (00): 1. European Commission. 0. Written declaration to the EU Court of Justice concerning case C- 1/. May. 0 European ch.indd 0 0//1 1:1:01