Introduction

In this chapter, we study the role of the EU and fellow member states in national policymaking during the refugee crisis. As we have already pointed out, the relationship between the EU and domestic politics has often been characterized as a two-level game. The two-level game concept is specifically related to international negotiations and captures the fact that international agreements have to be ratified at home. In the EU polity, however, the two-level game is not only or not even in the first place related to international negotiations. In the EU multilevel polity, the relationship between international and domestic politics is a two-way street, with international, that is, supra- and transnational, politics spilling over into domestic policymaking, and vice versa, domestic politics spilling over into EU policymaking. The interlocking of policymaking at the EU level with policymaking at the domestic level is particularly complex in a policy domain like asylum policy, where the EU and the member states share responsibility for policy. Moreover, such policymaking is complicated by the fact that the arena of cross-level policymaking in the EU is hardly structured by formal rules, which makes unilateral action by member states as likely as cooperative problem solving.Footnote ¹

For our analysis of this two-level game in the refugee crisis, we shall distinguish between two types of interactions between EU agencies and the member states, which we have already introduced in the theory chapter: “top-down” interventions, when EU policymaking or policymaking in other member states intervenes in domestic policies of a given member state, and “bottom-up” interventions, when domestic policymaking influences EU politics or the politics of other member states. In addition, we shall subdivide each type of intervention based on the prevailing conflict that has triggered it – an international (supranational or transnational) or a domestic conflict. Scholars of European integration have used the concept of “Europeanization” to assess the top-down effects of European interventions on domestic politics, that is, the “domestic adaptation to European integration” (Graziano and Vink Reference Green and Hobolt2006). This focus on top-down effects was a reaction to the long-term bottom-up focus on exploring the dynamics and potential outcomes of the European integration process (Börzel Reference Börzel2002: 193). Following Börzel (Reference Börzel2002), we propose to study here both the ways in which member state governments attempt to shape European policy outcomes and the ways in which they adapt to European policies. In contrast to our predecessors, however, we focus not on the eventual effects of Europeanization on national policy outcomesFootnote ² (although we come back to them in the conclusion of the chapter) but on the conflictual interactions between EU policymaking and policymaking in the member states.

First, we analyze the politicization of the forty national episodes in quantitative terms in order to show that episodes involving cross-level interventions are more highly politicized than purely domestic episodes. In a second step, we then choose episodes from four countries – Greece, Italy, Hungary, and Germany – to show in more detail how the cross-level interactions in the policymaking process operated during the refugee crisis.

A Typology of Cross-Level Interactions

Depending on the prevailing conflict, there are essentially two ways in which EU policymakers intervene in a top-down fashion in domestic politics. In the first way, there is a vertical conflict between the EU and a member state or a horizontal conflict between some member states with respect to the implementation of EU policy. The government of a given member state may fail to implement the joint EU policy, due to either lack of resources or lack of will. This is Börzel’s (Reference Börzel2002) case of “foot-dragging.” Such behavior by a member state may lead to attempts on the part of EU agencies to directly intervene in the implementation of EU policy at the domestic level. Domestic policymakers may welcome such interventions as they increase their capacity to act, but they are more likely to resist them because such interventions tend to come with strings attached. In the domain of asylum policy, as we have seen, the Dublin regulation places a major burden on frontline states for the implementation of the policy, and it is the frontline states that experienced difficulties in assuming their responsibilities during the refugee crisis. These difficulties led the destination states together with EU agencies to push for direct EU interventions in the frontline states – to improve their reception capacity (establishment of hotspots) and their capacity to patrol the external borders (upscaling Frontex into the EBCG) or to prevent secondary migration within the EU (as in transnational border conflicts between member states). Greece above all has been the object of EU interventions of this type.

In the second version of top-down interventions, it is the outcome of domestic policymaking that triggers an EU intervention into domestic politics. In this case, there is no question of foot-dragging with respect to EU policy – what is at stake here is the implementation of domestic policy that is the result of unilateral domestic policymaking and that is incompatible with or explicitly violates EU policy. In this case, the EU intervention is designed to prevent the unilateral domestic policy from being implemented. In the domain of asylum policy, this type of intervention has been applied to some of the policies regarding asylum rules adopted by Hungary because of their disregard for the rule of law.

Depending on the prevailing type of conflict, there are also two types of bottom-up interventions by member states in EU politics. The first version reminds us of Börzel’s “uploading” strategy, that is, a member state’s strategy of pushing a policy at the EU level that reflects the member state’s policy preferences and minimizes its implementation costs. Börzel conceived of this strategy, however, mainly in terms of regulation policies, while in the asylum policy domain during the refugee crisis, this kind of strategy applied above all to capacity building. According to this strategy, a member state unilaterally deals with an international challenge and adopts a policy that serves to substitute for the failure of the EU to adopt a joint policy to deal with the challenge in question. Bottom-up interventions by member state governments of this “self-help” type may be triggered by externalities created by a third country or by other member states. In the refugee crisis, this kind of intervention occurred in the case of frontline and transit states, which took a number of unilateral measures to police the external borders of the EU. Examples are the cases of Greece and Italy, which unilaterally had to deal with third countries – Turkey in the case of Greece and Libya in the case of Italy – in the absence of joint EU action. Hungary, too, built its own fences to unilaterally secure the external border of the EU, and Austria, in turn, organized the transnational cooperative effort to close the Balkan route as a substitute for the EU–Turkey agreement that had yet to materialize. Internal border closures can also be considered as examples of this type of bottom-up interventions to the extent that one member state unilaterally takes “rebordering” measures, that is, closes its borders with another member state and/or pushes back refugees coming from another member state.

In the second version of bottom-up interventions, domestic policymakers in some member states appeal to the EU and/or other member states to solve some domestic policy conflict. This appeal either calls for support in policy implementation (to alleviate the domestic burden) or attempts to signal that policy implementation at the domestic level is impossible because of too much domestic resistance. In the refugee crisis, it is the frontline and destination states that sought support for the redistribution of the refugees from the EU and the other member states. Germany above all sought the cooperation of its fellow member states for the accommodation of asylum seekers. Greece, as the frontline state most directly hit by refugee arrivals in summer and fall 2015, appealed to the EU for support to make up for its lack of capacity to deal with the inflow of refugees. The most conspicuous example of bottom-up signaling in reaction to EU measures during the refugee crisis is the Hungarian quota referendum, which was organized to send a message to the EU decision-makers that the EU’s relocation policy was incompatible with the situation in Hungarian domestic politics. Hungary’s use of domestic politics at the EU level most closely corresponds to what Putnam (Reference Putnam1988) originally had in mind with the two-level game concept: Weakness at home is a strength on the international stage. Domestic conflict implies the impossibility of a government cooperating internationally: Its hands are tied, and it cannot participate in joint solutions such as the redistribution of refugees across member states. Note that the domestic conflict, as in the case of the Hungarian quota referendum, may be deliberately created by the government of the member state for the purpose of strengthening its position in EU-level negotiations.

For the empirical classification of the national episodes into top-down and bottom-up types, we rely on the information about EU and member state actors targeting actors from the respective other level – domestic actors targeting international (supra- and transnational) actors and vice versa. Since such cross-level targeting is comparatively rare, we chose a low threshold to distinguish episodes with cross-level interactions from purely domestic episodes: If more than 20 percent of the actions in a given episode target actors from the respective other level, we classify it as a cross-level episode. Among the cross-level episodes thus identified, bottom-up targeting prevailed empirically. To qualify for the top-down types, at least 40 percent of the cross-level targeting actions had to be top-down. Based on prevailing conflict types, international or domestic conflicts (see Chapter 7), we then classify top-down and bottom-up episodes into their respective versions (see Table A11.1 in the appendix to this chapter for details).

Table 11.1 provides an overview over the four types of cross-level policy interactions and presents the episodes that are classified into the corresponding types. Six episodes (15 percent) are of the top-down type, four of the first variant, triggered by EU policies, and two of the second variant, triggered by domestic policymaking. Thirteen episodes (33 percent) represent bottom-up cross-level interactions, nine of which were triggered by EU policymaking and four by domestic policymaking. The remaining twenty-one episodes (53 percent) are of a purely domestic type.

Note that the distinction between cross-level and purely domestic episodes is closely related to the policy domain and to the type of conflict. Thus, all top-down episodes deal with border control issues, and all except one of the bottom-up episodes (the Greek International Protection Bill) also deal with border control issues or with relocation. By contrast, only five of the twenty-one domestic episodes are concerned with border control – the Italian Sicurezza laws, the Calais case (in both France and the UK this episode hardly involved European actors at all), Swedish border control, and the German suspension of the Dublin rules. In addition, two domestic episodes deal with resettlement/relocation – the British VPRS episode and the Swedish episode devoted to relocation between Swedish municipalities. Note that the very important case of the German suspension of the Dublin rules is misclassified by the rules applied here, that is, it is not classified as a cross-level episode. As we shall discuss in the next chapter in more detail, it is actually a case of a bottom-up cross-level episode, which we can see only when we link it systematically to the EU–Turkey agreement, an EU-level episode that was crucial for German policymaking during the crisis.

Cross-Level Politicization of Policymaking Episodes

This section presents a quantitative analysis of the politicization of the forty domestic policy episodes in order to show that cross-level episodes tend to be more highly politicized. The politicization of a policymaking episode is generally a function of exogenous and endogenous factors. Among the exogenous factors, as we have argued previously, the problem pressure and the political pressure exerted on the policymakers are crucial. The problem pressure is exogenous to the extent that the policymakers cannot influence the number of arrivals of refugees, at least not in the immediate term. The policy heritage – the combination of the responsibilities assigned to the member states by the prevailing EU policy and the limited resources available to come to terms with these responsibilities – is likely to restrict the options of the policymakers, especially in frontline and transit states. We expect the enormous problem pressure in these member states to contribute to the politicization of the policy episodes, independently of the political pressures.

The political pressure includes pressure from both domestic and international (supra- and transnational) politics. Domestic political pressure is likely to be endogenous to the domestic policymaking process. It may be driven by the national opposition, by domestic civil society actors, or by opposing factions within the country’s governing parties. Top-down international pressure by EU agencies and by other member states is a more exogenous factor that is likely to add and run counter to this domestic pressure. As Benz (Reference Bélanger and Meguid1992: 163f) has argued, linking domestic decision-making arenas with international arenas is likely to increase the conflict intensity of policymaking processes. Cross-level interactions introduce conflicts with supranational authorities and with other member states into domestic policymaking, which expands the scope of conflict and thus contributes to the politicization of national episodes. In particular, cross-level interactions may provide the government with an incentive to deliberately create domestic pressure to reinforce its position in the cross-level policymaking process. The politicization of domestic policymaking episodes by the national government may provide it with the reason for why it is unable to comply with EU policy requirements or for why it is required to unilaterally adopt policies that are incompatible with EU policies.

To test these expectations, we have created a dataset where the episode month constitutes the unit of analysis, that is, each episode is broken down into monthly units for which we calculate the level of politicization. The independent variables are the characteristics of the episode (cross-level interaction [top-down, bottom-up, or purely domestic] and conflict type [international or domestic]), type of member state, phase of the crisis (pre- and post-EU–Turkey agreement), and problem pressure. Table 11.2 presents the results of four increasingly complex regression models to explain the monthly politicization of the forty episodes. The first model includes only the characteristics of the episode, the second model adds the country type and the phase, the third model adds problem pressure, and the fourth model adds interactions between country type and phase.

Model 1 confirms the expectation that cross-level interactions increase the politicization of national policymaking episodes. Both top-down and bottom-up episodes are, on average, significantly more politicized than purely domestic episodes. Moreover, international conflicts are more highly politicized than domestic ones. The expansion of the scope of conflict beyond domestic politics apparently leads to an increase in politicization at the domestic level. Adding country type and phase in Model 2 doubles the R² from 0.10 to 0.18. Model 2 indicates that the politicization of the episodes has been greater in frontline states than in the other types of member states, a result that is attributable to the fact that all episodes in frontline states with the exception of one were characterized by cross-level interactions. Once we control for this effect, the effect of the cross-level interactions is considerably attenuated, and the effect of conflict type vanishes. The phase has, on average, no impact on politicization, which means that episodes before and after the adoption of the EU–Turkey agreement were equally politicized.

Model 3 adds our indicator for problem pressure, which has a highly significant effect on politicization, independently of the effects of the indicators already included in Model 2. Adding problem pressure, however, hardly modifies the effects of the indicators previously introduced, which is to suggest that the greater politicization of the episodes in frontline states is attributable not only to problem pressure but also to some extent to endogenous political pressure. Nor does adding problem pressure modify the R². Model 4 specifies that the increased politicization in the frontline states occurs mainly in the second phase, that is, after the peak of the crisis when the immediate problem pressure has become less pronounced. This is yet another indication that the politicization of the crisis in the frontline states was, to some extent at least, the endogenous result of domestic politics and only partly the result of exogenous problem pressure. Once we take the endogenous politicization in the frontline states into account, the effect of the two cross-level interactions on the politicization of the episodes is again significantly enhanced. In other words, cross-level policymaking increases the politicization of the episodes in general and is not a specialty of the frontline states.

Having clarified this general point, we now turn to a detailed analysis of the variety of cross-level policymaking in the four member states, where it was most important during the refugee crisis. The Greek case will serve to illustrate both EU policy triggering top-down EU interventions to increase the domestic capacity of a frontline state to deal with the crisis and bottom-up demands of a frontline state for EU support. The case of Italy, our second frontline state, will focus on bottom-up efforts to substitute unilaterally for EU policy but will also feature an episode of top-down intervention by the EU to come to terms with externalities created by Italian policy for its neighbors. In contrast to the Greek case, the Italian example will show how factors endogenous to domestic policymaking are creating international conflicts and cross-level interactions. Third, the Hungarian case will above all serve to discuss top-down and bottom-up cross-level interactions that are rooted in conflicts endogenously created in domestic politics. Finally, the German episode will show how domestic policymaking in a member state can trigger EU policymaking in support of the member state.

Greece: The Frontline State Facing the Most Conspicuous International Interventions

Greece is the member state where intervention in domestic politics by EU agencies and governments from a third country (Turkey) and from other member states were most conspicuous. All five Greek episodes are characterized by international conflicts, which are associated either with top-down interventions in domestic politics or with bottom-up interventions of Greece at the EU level. Moreover, all Greek episodes respond to extraordinary problem pressure, given that Greece was the member states where the arrivals of refugees were concentrated, both in phase 1 and at the end of phase 2 of the crisis.

Phase 1: Summer 2015 and Hotspots

At the peak of the refugee crisis, the politicization of asylum policymaking in Greece was closely aligned with the politicization of the crisis at the EU level, as is shown by the left-hand graph in the first row of Figure 11.1, which presents the politicization of the Greek episodes by phase and adds the politicization of the EU episodes (mostly focusing on Hotspots, EBCG, Relocation and the EU–Turkey agreement) in phase 1. The negotiations related to the EU–Turkey agreement, and even more directly the elaboration of the hotspot approach and the transformation of Frontex into the EBCG, were of immediate concern to Greece. Accordingly, Greek policymaking in summer and fall 2015 and in early 2016 took place in the shadow of EU policymaking.

Figure 11.1 Politicization of Greek episodes

In summer 2015, when Greece was first hit by the flood of refugee arrivals, the country was in fact preoccupied with the bailout process and not properly equipped and hardly willing to deal with the inflow of refugees (see Chapter 4). As is argued by Nestoras (Reference Netjes and Binnema2015: 19), the “intention to use the migration crisis in order to leverage some form of financial relief – extra funds or relaxed bailout terms – or simply to claim a moral high ground was evident from the beginning of Syriza’s term in power.”Footnote ³ There was “an explicit attempt to connect the Euro-crisis with the migration crisis and bargain with Greece’s position as a gateway to Europe” (p. 20). Nestoras cites Defense Minister Panos Kammenos, the leader of Syriza’s far right coalition partner, who did not hold back when he threatened (in March 2015) to send migrants, including jihadists, to western Europe: “If Europe leaves us in the crisis, we will flood it with migrants.”Footnote ⁴ In the summer 2015 episode, the Greek government appealed to the EU for funds to manage the refugees. But once the EU promised to deliver, Greece was unable to administer the promised funds. Only in mid-September did the Commission announce that it had received all the required documentation regarding the management of these funds and promised to process it as quickly as possible to release the first 30 million euros (of a total allotted sum of more than 500 million euros). The Greek ministers and deputy ministers responsible for migration and foreign affairs multiplied the declaratory statements and symbolic gestures, as did the EU commissioner for interior affairs and migration, Demetri Avramopoulos, a Greek, as well as an assortment of government and EU spokespersons. But nothing much actually happened. The Greek opposition was asking for the resignation of the minister of migration policy, criticizing the government for “deafening inaction” and “complete absence of a plan.” And on August 28, in the midst of the crisis, the whole government did, indeed, resign – but for reasons having to do with the bailout, not with the refugee crisis.

As is shown by the right-hand graph in the top row of Figure 11.1, this first episode was immediately followed by the more intensely politicized episode of the hotspots. The latter episode is both an EU-level episode and a Greek episode, and it represents the most clear-cut case of a top-down intervention of EU agencies and fellow member states in the domestic policymaking of a member state. What we present here is the politicization of the episode in Greece, which covers the period from October 2015, when the first deal to implement the hotspot policy was struck, to May 2016, when the final makeshift migrant camps were evacuated and the hotspot approach was fully implemented. The two graphs in the bottom row of Figure 11.1 document that during this period, domestic and cross-level politicization developed in lock-step, reaching comparable levels. The same applies to top-down and bottom-up cross-level politicization.

At the end of summer 2015, domestic politics loomed large in Greece as the country was preparing for new elections, which were to take place on September 20. Moreover, domestic politics were still dominated by the issue of the bailout and the memorandum process. With the preceding government having resigned, it was up to the Greek president to perform the symbolic gestures in asylum policy during the interregnum. The new government, which was practically the same as the old one, took office immediately after the elections, at a moment when the European governments were in the thrall of the relocation issue, which they tried to resolve under German pressure. Under the pressure of the events, the exchanges between the new government and European officials, presidents, prime ministers, and ministers of other EU member states became ever more intense, not only at European summits but also in bilateral meetings on the phone and in person. European worthies came to visit Greece to inspect the sites and to get an idea of the proportions of the problem, while Greek officials intervened with the Commission and fellow ministers in other member states to explain the Greek predicament. The EU expected Greece to set up hotspots and promised its help in setting them up, but Greece was reluctant to do so because it was afraid that the hotspots would be perceived as an alternative to relocations. Several ministers proposed that an alternative could be to build the hotspots and refugee centers directly in Turkey. Of course, when Prime Minister Alexis Tsipras went to Ankara to explore the prospects of such a proposal, he found that the Turkish prime minister, Davutoglu, was afraid of the exact same trap, that is, that relocations would never happen once hotspots were set up, and was therefore similarly reluctant to construct them in Turkey, which meant that the hot potato returned again to Greece.

As time passed, the pressure on the Greek government to get things done – to construct the hotspots and to stop the inflow – increased. The Greek strategy of evading the issue – an example of Börzel’s “foot-dragging” – proved to be increasingly vulnerable to the demand from other member states to exclude it from the Schengen area and to the determination of the Balkan countries to shut down their borders. Demands from the V4 countries for the removal of Greece from the Schengen area in December 2015 provoked a reply from the Greek minister of foreign affairs, who pointed out that the dimensions of the problem were bigger than any country of any size could handle and that it was unreasonable to expect a national solution from Greece for the joint problem. Greek protests notwithstanding, by the end of November 2015, the North Macedonian government started putting up a fence and sent police to the border, blocking the continuation of the flows along the Balkan route. Concurrently, the European institutional pressure on Greece to conform increased, and threats of excluding it temporarily from Schengen persisted. Eventually, at the end of January, the Commission gave Greece a three-month “warning” to fix the issues with border control and registration, or a temporary suspension from Schengen would be imposed. Moreover, at the West Balkan conference at the end of February 2016, under the leadership of Austria, the western Balkan countries – four EU member states (Austria, Bulgaria, Croatia, and Slovenia) and six candidate countries from the western Balkans (Albania, Bosnia-Herzegovina, Kosovo, Macedonia, Montenegro, and Serbia) – agreed to shut down their borders. They started to do so immediately after the conference.

Faced with this threat, in January 2016, Greece joined Germany in its efforts to come to an agreement with Turkey. At this point, Prime Minister Tsipras explicitly stated that the key to the refugee crisis was “transferring the focus of the refugee crisis management to Turkey.” By early February, the Greek government admitted Turkey to its list of “safe countries,” so the returns to Turkey could be legally unblocked. By mid-February 2016, Germany and the key EU actors, in turn, took the side of Greece in its struggle with the Balkan countries, with Chancellor Merkel, EU Council president Tusk, and EU Commission president Juncker declaring over a succession of days that Greece could not be left to fend for itself and that the solution of closing the Balkan route was not really a solution. The EU assumed a mediating role between the two “blocks” of member states that faced each other at this point, the “hardline” transit and bystander states led by Austria, which wanted Greece to control its borders or be expelled from Schengen, and the more moderate western destination states, like Germany and France, which were more focused on the maintenance of Schengen.

With respect to the hotspots, the Greek government ended up taking some necessary steps. At the summit in mid-December, Prime Minister Tsipras assured German chancellor Merkel that the hotspots would be completed within the next two months. To this end, the government mobilized the army to speed up their construction at the end of January 2016. Conscripts serving on Lesvos and the other islands assisted in the construction. Locals protested and blockaded the hotspot installations, but the mainstream opposition chose to mostly leave the refugee crisis outside of domestic political conflict. Eventually, the episode became a race to the finish line, to halt the refugee flows and set up the hotspots before Greece was expelled de facto or de jure from the Schengen zone. The EU–Turkey agreement was a huge relief to the tension, as was completion of the hotspots, owing much to the army’s assistance. It is not clear whether the hotspots and Frontex’s assistance would have been able to stem the tide of refugees without the agreement with Turkey that ground arrivals to a halt. The episode formally ended with the disbandment of the camp at Idomeni on the Macedonian border in late May 2016.

Italy: A Frontline State Substituting Domestic Policy for Joint EU Solutions

Italy is the other frontline state in our country selection – a frontline state that was, however, much less affected by the refugee crisis of 2015–16 than Greece was. Four of the five Italian episodes concern cross-level interactions, mainly of the bottom-up type with prevailing international conflicts – the Mare Nostrum, Brenner, and Ventimiglia episodes during the first phase and the episode of Port Closures during the second phase. To these episodes should be added the EU–Libya conflict, an EU-level episode that was actually initiated by unilateral policy measures on the part of Italy. As is shown in the left-hand graph of Figure 11.2, the Mare Nostrum episode and one of the border disputes (Ventimiglia) preceded the peak of the refugee crisis, while the other border dispute (Brenner) took place at the peak of the crisis. The most politicized episode, however, about Port Closures, occurred late in the second phase and was almost entirely unconnected to problem pressure exerted by the crisis. The right-hand graph of Figure 11.2 indicates that these episodes also gave rise to domestic politicization, but international politicization prevailed, except for the very last episode, the purely domestic episode concerning the Sicurezza decrees.

Figure 11.2 Politicization of Italian episodes

Hungary: A Variety of Cross-Level Interactions Rooted in Domestic Conflicts

All five Hungarian episodes involve cross-level interactions, though they were of varying types. Two episodes – the Fence Building and the Legal Border Barrier Amendment – refer to unilateral actions by Hungary to substitute for joint EU measures to protect the external border. Two episodes – the Civil Law of 2017 imposing a financial disclosure requirement on all NGOs receiving funding from abroad and the “Stop Soros” package of 2018 imposing an even more onerous special “migration tax” on all organizations deemed to aid immigrants – are domestic measures in Hungary that led to EU interventions to rectify domestic policy. The fifth episode – the quota referendum of 2016, the Hungarian response to the European attempt to introduce a relocation scheme – represents the case of a domestic policy signaling to the EU and the other member states domestic obstacles to the implementation of EU policy. The quota referendum was the most politicized Hungarian episode and the most politicized of all national episodes. Four of the five Hungarian episodes were highly politicized, even when compared to the high level of politicization of episodes in frontline states (see Table 5.2).

Figure 11.3 presents the politicization of the Hungarian episodes. The left-hand graph compares the border control episodes (Fence Building and Legal Border Barrier Amendment) with the episodes addressing asylum rules (Quota Referendum, Civil Law, and Stop Soros) and with the EU episodes addressing asylum rules (Relocation and Dublin Reform). As we can see, at first, the Hungarian politicization of border control moves in parallel with the politicization of asylum rules at the EU level. The two developments, however, part ways as the crisis starts in earnest. Moreover, the politicization of asylum rules at the domestic level is completely uncoupled from the corresponding politicization at the EU level. It unfolds in three waves that correspond to the three episodes dealing with relocation quotas, Civil Law, and Soros. The politicization of the Hungarian asylum rules proves to have been much more intense than the politicization of these rules at the EU level and also more intense than the politicization of border controls, except for the very beginning of the crisis, when Hungary started with its fence building. Contrary to what we have observed in the frontline states, the Hungarian politicization essentially follows a domestic logic, as is illustrated by the right-hand graph of Figure 11.3: Throughout the crisis, the domestic politicization has been more intense than the cross-level politicization. We focus here on the bottom-up episodes, since the Hungarian top-down episodes (Fence Building and the Legal Border Barrier Amendment) have already been discussed in some detail in Chapter 6.

Figure 11.3 Politicization of the episodes in Hungary

Germany: Seeking EU Support to Overcome Domestic Conflicts

Germany provides two episodes of domestic, intragovernmental conflicts that led the government to seek support from the EU to solve the domestic conflicts. The first example concerns the episode of the suspension of the Dublin regulation by Germany in September 2015. This episode is classified as a purely domestic episode, which is misleading because it is intimately linked to the EU–Turkey agreement, which served as the German chancellor’s plan B to come to terms with the domestic conflict that had been unleashed by her unprecedented decision to suspend the Dublin regulation and to admit refugees to Germany who had traveled from Greece across the Balkan route to ask for asylum in Germany. The episode of the EU–Turkey agreement is an EU-level episode – but one that was intensely discussed in Germany. According to the criteria applied to classify cross-level episodes, the German discussion of this agreement would qualify as a bottom-up cross-level episode: More than 40 percent of the actions reported in the German debate on the EU–Turkey agreement involved cross-level interactions, and the overwhelming majority of these cross-level interactions were of the bottom-up type. We shall discuss this episode in more detail in the next chapter. The other example of intragovernmental German conflicts spilling over to the European level is the CDU-CSU Conflict in summer 2018, which also induced the German chancellor to seek support at the EU level to solve her differences with her coalition partners. This episode qualifies as a bottom-up cross-level episode rooted in domestic conflicts.

Figure 11.4 shows the close alignment of German domestic politicization with the politicization of asylum rules (Relocation Quota and Dublin Reform) at the EU level during the first phase and then again in summer 2018. This alignment is a result of spillover processes from German policymaking to the EU level. In the first phase, as Germany attempted to come to terms with the crisis domestically, it at the same time put pressure on the other member states to get the relocation quota passed in the Council of Ministers in an attempt to share the burden of reception and integration of asylum seekers. It is only once Germany failed to obtain a relocation scheme from its fellow member states that it turned to an agreement with Turkey as the second best solution. The renewed alignment of German policymaking with the politicization of asylum rules at the EU level in summer 2018 is the result of yet another spillover of domestic German conflicts to the EU level. In both instances, it was mainly intragovernmental conflicts that led to the cross-level politicization of policymaking.

Figure 11.4 Politicization of German episodes and EU episodes concerning asylum rules

The border control issue returned to German politics when Horst Seehofer, the head of the CSU and the most vocal critic of Merkel’s open-doors approach in 2015–16, became minister of the interior in Merkel’s new grand coalition cabinet that took office in March 2018. It was Seehofer’s attempt to implement his hardline asylum policy that gave rise to the second border control episode in Germany. In early June 2018, Seehofer insisted on turning back at the German border two categories of refugees: those who had already been registered in other countries and those against whom a reentry ban had been imposed in the past. He met with resistance on the part of Chancellor Merkel, who had legal and practical objections and pleaded for a coordinated European solution instead. The issue unleashed an open power struggle between the two, which developed into a highly politicized episode (although it does not register as such in Table 5.2, because of its very short duration).

On June 18, 2018, Merkel asked Seehofer for a two-week timeout to solve the issue at the European level. More specifically, Merkel wanted to negotiate bilateral return agreements with Italy and Greece so that refugees could be returned in a coordinated manner, plus a “European solution” that she promised to offer as an alternative to Seehofer’s approach involving rejections at the border. A week and a half before the upcoming European summit, there was, however, little clarity about what such a “European solution” would look like. Merkel intensified cross-national negotiations in preparation for the upcoming summit. First, she seized upon the occasion of the Franco–German summit at Merseburg Castle on June 19 to discuss curbing migration with French president Emmanuel Macron. Macron assured Chancellor Merkel (CDU) of his support to find, “together with some other states,” solutions to sending back already registered refugees. Macron promised to speak to Italy’s new prime minister Giuseppe Conte, who had just taken office as the head of the Lega–M5S coalition government. Next, she relied on EU Commission president Jean-Claude Juncker for the organization of a preparatory summit of “interested states” in the run-up to the European summit of June 28–29. At the request of Merkel, Juncker invited the heads of state and government of sixteen particularly affected EU countries (among them Austria, Italy, France, Greece, Bulgaria, and Spain) to Brussels for a meeting a week before the summit of the European Council to discuss a “European solution” to the migration crisis.

Meanwhile, Seehofer insisted on sending a signal to the German public. He ordered that the federal police should, starting on July 1, reject refugees against whom a reentry ban had been imposed in the past, and he reiterated the proposal to reject refugees who had already been registered for asylum in another EU country. He threatened to break up the coalition if his plan were not adopted, and CSU parliamentary group leader Alexander Dobrindt no longer ruled out that the dispute over the refugee policy could mean the end of the union party comprised of CDU and CSU. The SPD, in turn, was urging the coalition partners CDU and CSU to resolve their asylum dispute before the next coalition committee meeting. Federal president Steinmeier (SPD) heavily criticized the conflict between CDU and CSU and supported Merkel’s plea for a joint EU solution to the conflict about the reform of the European migration policy.

In the government declaration in the Bundestag just before the European summit, Chancellor Merkel (CDU) spoke engagingly. She warned against a unilateral German solution and suggested that asylum policy could become a fateful issue for the future of Europe. Seehofer was not present in the plenum, and the CSU reacted coolly. At the summit in Brussels, Merkel fought for her job. At first, a compromise failed to materialize. Merkel met with massive resistance from Italian prime minister Giuseppe Conte, who blocked all decisions that had been prepared in the run-up to the summit. He asked for a radically new policy, which would include abandoning the Dublin rule. Although Germany and Italy shared common interests as key frontline and destination states, they failed to find a common ground at the meeting. While Conte was ready to understand the asylum issue as one concerning the whole of Europe, he refused to accept that the obligation to rescue people at sea implied the obligation to treat their asylum requests in the name of all of Europe. EU Council president Donald Tusk and EU Commission president Jean-Claude Juncker were then forced to cancel their scheduled press conference. Eventually, however, the heads of state arrived at an agreement: A concept for disembarkation platforms would have to be elaborated for refugees who had been rescued in the Mediterranean – this was a measure to reduce the attractiveness of the business model of smugglers (in response to an old Italian demand and building on an idea of Tusk). In addition, so-called controlled centers were to be built by member states on a voluntary basis, where decisions would be taken about who had a right to protection (an idea of Macron). The recognized refugees would be distributed over the member states – on a voluntary basis. However, much was still unclear about details and, as it turned out, the proposals remained a dead letter.

Merkel, however, was relieved. She had achieved little in terms of a solution to the migration crisis but a lot in terms of saving her chancellorship. At the press conference following the summit, she was asked whether the result of the summit was functionally equivalent to the immediate rejection of already registered refugees at the border (as demanded by Seehofer and the CSU). She claimed that if everything were to be implemented as discussed, the adopted proposal would be more than functionally equivalent and there would be real progress. Seehofer insisted that the summit solution was not functionally equivalent, but, surprisingly, in a direct meeting between the two on July 2, he and Merkel arrived at a compromise. Refugees who were caught at the border, although not allowed to enter or stay in Germany because they had been already rejected previously (a very small group indeed), were exempted from the compromise because Seehofer had already ordered the federal police to reject them at the border after July 1. Refugees who had already been registered in another country where they had asked for asylum (a larger, but still comparatively small, group of 35,000 persons per year) would be directly returned to the country responsible for them – but only if there was an agreement with the country in question. If there was no such agreement, they would be rejected at the border with Austria. Those refugees who were rejected were to be put into buildings of the federal police close to the border or in the transit zone of the Munich airport (the so-called transit centers).

Nobody knew exactly what the compromise implied in practice and whether it was legally possible to implement it. The SPD angrily opposed the transit centers, and the opposition voiced a sharp critique. Austrian prime minister Sebastian Kurz, caught off guard by this asylum compromise, issued a sharp reaction: “We are certainly not ready to conclude contracts at the expense of Austria.” In a subsequent joint meeting, Seehofer and Kurz decided to increase the pressure on the Italian government to take back refugees who desired to go to Germany. Meeting shortly afterward, the interior ministers of Germany, Austria, and Italy tried to negotiate an agreement about the return of asylum seekers to Italy. Meanwhile, Merkel tried to accommodate the SPD, declaring that under the German constitution, asylum seekers could be held in transit centers for a maximum of two days. If the transfer to the country where they had already been registered was not successful within this lapse of time, they would have to be brought to regular facilities. Nevertheless, Seehofer considered his conflict with Chancellor Merkel about the refugee policy to be over: There were disagreements about content but no personal bad feelings, he claimed. They could “look each other in the eye” even after an argument. Seehofer justified his threat of resignation by claiming that he would not allow himself to be thrown out of office by a chancellor “who was Chancellor only because of me.” On July 10, he finally presented his “master plan” for migration policy for faster asylum procedures and more consistent deportations, which he had already announced in March, even before the new government was sworn in, but was prevented from publishing by the conflict with the chancellor. Facilitated by the EU-level interlude, the compromise in early July essentially served as a face-saving device for Merkel and Seehofer and did not change much in Germany’s asylum policy.

Conclusion

In this chapter, we took a closer look at the cross-level episodes, which include roughly half of the national episodes of our study. This is a remarkably high share, which indicates that national asylum policymaking is taking place in the shadow of EU policymaking. These episodes have been more intensely politicized than purely domestic episodes, since they involved the expansion of conflict beyond the national borders both in a transnational and in a vertical direction. Cross-level episodes have either been rooted in domestic conflicts that expanded up into the international realm or in international conflicts that were closely associated with domestic politics. We have presented a fourfold typology of such cross-level episodes, which distinguishes between top-down and bottom-up cross-level interventions for both international and domestic conflicts. Top-down interventions involve attempts of EU agencies and/or fellow member states to impose EU policy implementation on a defaulting member state – either by providing support or by imposing disciplinary measures – or to prevent a member state from implementing domestic policies that are incompatible with fundamental EU values. Support may be forthcoming in terms of capacity building (providing the member state with additional resources), in terms of regulation (adapting some policies to the needs of the member state), or in more exclusively symbolic terms. As we have seen, additional resources have been provided to Greece in the hotspot episode, and to Italy in the context of Triton and the EU–Libya agreement. Support has also been pledged in more symbolic terms, as in the case of Greece’s border conflict with Turkey and in the case of the German intragovernmental conflict in 2018. However, in regulatory terms, support for frontline and destination states has not been forthcoming, and several of the episodes just ended nowhere, with the attention of the public and policymakers turning elsewhere and leaving the issue lingering. Calling a defaulting member state back to order may include material sanctions but also punishments such as exclusion, shaming, and shunning, as is illustrated by the Hungarian Civil Law and “Stop Soros” episodes. In the refugee crisis, such measures have been ineffective.

Bottom-up interventions involve unilateral policy measures on the part of a member state to substitute for EU policies that have not been forthcoming, the appeal by a member state to the EU/fellow member states for help, or its signaling of the impossibility of implementing joint policies. Faced with unilateral measures by member states, the EU/fellow member states may attempt to mediate between the member state adopting the measure and other member states directly concerned by the externalities of the measure, as has occurred in several of the cases we have reviewed here (the border conflicts between member states, the Italian Port Closures). The EU may also attempt to develop a policy of its own that is able to build on and replace the unilateral policy of the member state in question, as in the cases of Mare Nostrum, the EU–Libya agreement, and the EU–Turkey agreement (which served to replace the unilateral Balkan Route Closure). But the unilateral action by a member state may also prevent the EU from adopting joint solutions and have a paralyzing effect, like the Hungarian quota referendum and the associated actions of the V4.

The intense cross-level interactions in the domestic episodes during the refugee crisis demonstrate the interdependencies between the member states and between the member states and the EU in this policy domain. At the same time, they also demonstrate the difficulties in coming to joint solutions, even under great pressure, and the amount of effort that it takes to search for joint policies in a polity that requires consensual decision-making.