Introduction

Are immigration crises in the European Union (EU) differently politicized than economic crises? If so, in what way? This article investigates the generalizability of democratic functionalism as a theoretical framework explaining mechanisms of politicization beyond the Euro crisis, on which the literature originally focused (Statham and Trenz 2015). The goal is to show the theory’s limited applicability to earlier episodes of immigration crises in the EU—before the 2015 episode which generated high political salience and strong contestation from the public in several member states (Debomy and Tripier 2017; Schimmelfennig 2017, p. 8). In contrast to the Euro crisis, where leaders’ depoliticization attempts backfired and opened a window of opportunity for fringe parties to mobilize public opinion against the EU more broadly, the case examined in this article shows that EU governments themselves can instrumentalize the media during immigration crises, without political mobilization from the public. Moreover, the polarization of public opinion during immigration crises occurs along exclusionary national identities rather than pro-/anti-European lines—as seen during the Euro crisis.

The empirical analysis focuses on a single case study illustrating what is generally considered the most politicized issue in the area of border management in the post-Lisbon era—prior to the 2015 crisis—namely whether to lawfully allow the temporary reintroduction of internal border controls in the Schengen Area. Known in the legislative process as the ‘Schengen Governance Package’ (2011–2013), the case originated in a crisis situation caused by an inflow of Tunisian refugees fleeing from the Arab Spring and trying to reach France by transiting Italy (Pascouau 2011). The episode led to a politicized dispute between French and Italian authorities that snowballed at the EU level and became a bone of contention between EU institutions involved in decision-making over the Schengen regime. Given the multiple levels of clashes reflected in the file, the case offers an excellent basis for examining ‘mechanisms of politicization’ from a democratic functionalist lens.

Democratic functionalism is a theoretical framework originally designed to capture the democratic character of the EU (Trenz and Eder 2004) and later developed to explain the phenomenon of ‘EU politicization’—defined as the contestation of EU issues in national public spheres as reflected in media debates (Statham and Trenz 2013). In the context of the Euro crisis, the theory was further refined to explain how politicization occurs by specifying the ‘mechanisms of EU politicization’ (Statham and Trenz 2015). The focus on the Euro crisis was shared with similar research on the politicization of crisis situations in the EU (Kriesi and Grande 2016; Leupold 2016; Schimmelfennig 2014; Wonka 2016). In contrast, comparisons between economic and immigration crises have only recently started to emerge, addressing the implications of the crises on European party systems (Hooghe and Marks 2017) or explaining EU responses to the crises through the lens of integration theories (Börzel and Risse 2017; Genschel and Jachtenfuchs 2017; Schimmelfennig 2017). What remains under-researched are the mechanisms of politicization during immigration crises—understood in the EU setting as the occurrence of migratory pressures and/or migrant deaths at the external borders of the Union, often caused by waves of refugees.

While immigration is a topic generally politicized at the national level, the EU has the additional dimension of ‘Schengen’—the area of 26 European countries which allow free movement of people—whose very existence was threatened during the refugee crisis of 2015 (Binyon 2015). Indeed, the cracks in the EU’s asylum system were publicly revealed when over a million refugees fleeing from war in Syria and Iraq crossed the Mediterranean and took the so-called Balkan route in order to reach asylum-friendly member states like Germany and Sweden (Trauner 2016, p. 311). According to Eurobarometer surveys, immigration soon became the top concern of EU citizens irrespective of the impact of the refugee crisis on their member state, e.g. respondents in Estonia, Malta, Hungary, and the Czech Republic perceived immigration in a similar way despite being unevenly affected by the crisis (Debomy and Tripier 2017, pp. 16–17). Anti-immigrant sentiment simultaneously led to the increase in the vote share of right-wing populist parties in Germany, France, the Netherlands, Hungary, Greece, Austria, Sweden, Denmark, and Slovakia (The Guardian 2016). Most infamously, Hungarian Prime Minister Viktor Orbán built a border fence to stem the refugee inflow to his country, rejected the EU refugee relocation quotas, and even organized a referendum on the matter—supported by 98% of the 40.4% electorate who participated (BBC News 2016). But although the magnitude of the 2015 refugee inflow was unparalleled, the EU has been confronted with other immigration crises since the establishment of the Schengen Area in 1995 (European Commission, n.d.). The question of how earlier crisis episodes differ in terms of the mechanisms of politicization lies at the centre of this article.

The remainder of the article is structured as follows. The first part provides an overview of the theoretical application of democratic functionalism to crisis governance in the EU, starting from a general definition of politicization and ending with an operationalization of the concept. The second part puts forth the case study selected, discussing its relevance, the general lines of politicization identified thereof, as well as the dimensions of (1) polarization and (2) public resonance in relation to (3) legitimation. The conclusion problematizes the recent refugee crisis from the perspective of democratic functionalism, emphasizing the differences from the episode analysed in this article. Recent events suggest that the flaws of democratic functionalism could be overcome if the theory develops an account of how much and for how long an EU issue needs to be contested in the public sphere for the ‘mechanisms of politicization’ to be visible.

Democratic functionalism, crisis management, and the (re-) politicization of issues

The progressive politicization of EU governance starting the early 1990s is an established area of research in EU studies. Under the theoretical label of post-functionalism, Hooghe and Marks were the first to distinguish this development from previous periods in European integration history, when elite decision-making over European issues remained insulated from domestic public opinion (Hooghe and Marks 2009, p. 5). As ‘politicization’ gained currency in academic circles, so did the inherent ambiguity of the term. The work of de Wilde and associates, on the one hand, and Grande and Hutter, on the other hand, contributed to a more precise conceptualization of EU politicization. They defined the term as a process characterized by (a) the increased salience of EU governance in public debates, (b) the polarization of opinions expressed about EU policy decisions, and (c) the expansion of actors and audiences involved in these public debates (De Wilde et al. 2016, p. 4; Grande and Hutter 2016, p. 25). Focused on national arenas, their line of research found so far that EU politicization is ‘differentiated […] across times, countries and settings’ (De Wilde et al. 2016, p. 3).

Nowhere is the presence of politicization more controversial than in relation to Euro crisis management. In this context, a fascinating academic debate took shape around the sovereign debt crisis of Greece and the economic reforms consequently adopted at the EU level. In an article claiming the limited explanatory power of post-functionalism to account for EU responses to the Euro crisis, Frank Schimmelfennig argued that ‘governments have learned to contain the constraining dissensus by isolating crisis management from politicization’ (Schimmelfennig 2014, p. 331). The idea was that even though European integration issues have become subject since the Maastricht Treaty to preferences of ‘mass publics’ expressed in national elections or referendums, this has not impacted EU decision-making ‘when it counts’, namely in times of crisis. Conversely, Schimmelfennig maintained that political élites found ways to bypass public opinion during crisis situations, for example, by avoiding plebiscites on the second Greek bailout (2011) or for the ratification of the Fiscal Compact (2012). This argument held until 2015, when the Greek referendum on a new bailout package and the ensuing negotiations which required parliamentary ratification from seven other Eurozone members (The Guardian 2015) showed that EU decision-making can conversely become heavily politicized in times of crisis. The mechanisms through which this process of politicization happened, nonetheless, were only specified within the framework of democratic functionalism, whose main tenets are discussed next.

Statham and Trenz, the proponents of democratic functionalism, have a long-standing interest in the empirical manifestations of politicization in Europe, having earlier conducted a comprehensive study on the public contestation of the failed Constitutional Treaty (Statham and Trenz 2013). In what they call the public sphereFootnote 1 perspective, politicization is conceptualized as a particular type of political conflict whose distinctiveness is determined by its public character as reflected in the mass media, which allows political actors to make competitive claims that resonate with (segments of) the public. In the EU context, this lens implies looking at how European issues are contentiously debated in national public spheres, expected to become Europeanized because of their constant mentioning of actors and decisions taken at the EU level or in other member states. The media constitutes the environment through which the expanding public discourse is brought to the audience, but the process is not unidirectional; on the contrary, there are feedback loops that make political actors aware of the resonance of their debates for the public. From this standpoint, a public sphere is expected to ‘fulfil an important democratizing function: it enhances legitimacy by making executive decisions transparent, including civil society, and provides important critical feedback to policy decisions’ (Statham and Trenz 2013, p. 7). The understanding of ‘democracy’ here is limited and follows the normative, process-oriented logic according to which the more the EU is publicly criticized for its democratic deficit, the more democratic its political institutions become (Trenz and Eder 2004, p. 7).

Within the framework of this public sphere approach, Statham and Trenz theorize how politicization takes place in the EU (Statham and Trenz 2015, p. 287). Three mechanisms are identified, revolving around ‘the type of conflict that produces [politicization], the public arena in which it unfolds and its potential legitimating or delegitimating effects for the polity’ (ibid., p. 294). In brief, a European issue is expected to become politicized (1) if it is subject to polarization between pro- and anti-EU camps, (2) when it is picked up by the media, without which the aforementioned polarization would remain invisible; and (3) when the above-mentioned camps make competitive public claims aimed at legitimation—meaning that by attempting to demonstrate the merits of their views, political actors essentially bring into public contestation the very legitimacy of the European political order (ibid, pp. 293–6). From an analytical standpoint, the third mechanism remains underspecified, as the authors refer to legitimacy contestation and the legitimacy of the EU political system interchangeably. This approach is consistent, however, with the normative perspective described earlier, namely the more you contest something, the more legitimate it becomes.

Furthermore, the crisis context is hypothesized to increase the likelihood of all three mechanisms. With regard to the Euro crisis in particular, Statham and Trenz argue that it was precisely the depoliticization attempts of national governments that heightened the cleavage between pro- and anti-integrationist camps by creating discursive opportunities for fringe parties and social movements to contest EU responses to crisis; in other words, the absence of élites from the public sphere was substituted by new actors who disagreed with the mainstream and thus produced a ‘re-politicization’ of issues (ibid., p. 297). This would not have been possible, however, without the close media attention devoted to the crisis—a reliable generator of ‘breaking news’ which came to dominate tabloid publications and social media, therefore strengthening the role of mass media more generally ‘as the amplifier of popular voice’ (ibid., p. 298). Under the circumstances, the initial polarization concerning potential solutions to the crisis became a debate about the ‘political legitimacy of both the nation state and the EU’ (ibid., p. 301)—as revealed in the summer of 2015 by the politicization of the new Greek bailout. But if this account fits closely the unfolding of events in the Euro crisis, the question remains whether the theoretical tenets of democratic functionalism travel to different contexts, i.e. crisis situations in other policy areas.

Before moving to a discussion of politicization in relation to immigration crises, it is essential to clarify the methodological approach to politicization adopted for the purposes of empirical investigation. Statham and Trenz follow a similar understanding (Statham and Trenz 2015, p. 304), as they agree to operationalize politicization as ‘the practice of competitive representative claims-making in the public sphere’—observed when officials make public contradicting claims to represent a particular set of interests or values presumably held by their target constituency (De Wilde 2011, p. 572). Using an adapted version of the method of ‘claims analysis’ originally developed by Koopmans and Statham in the context of the social movements literature (1999), this approach takes a constructivist view on representation based on the political theory work of Michael Saward on the ‘representative claim’. Herein, representation is first and foremost understood as a performative act in which political actors construct representative links between themselves and their constituency and thus make them real, not least in order to address their ‘sense of remoteness and inadequacy’ (Saward 2006, pp. 299–301; emphasis in original).

The principal merit of Saward’s approach comes from deconstructing the process of representative claims-making into its parts. The starting point is the maker of representative claims, who says something about a subject (the representative) which claims to represent an object (the constituency) (ibid., pp. 301–2). To be politically acceptable, representative claims draw on familiar terms and understandings (the referent), ‘ready-mades’ which potential audiences can first recognize and then accept or reject (ibid., p. 303). The division into five components does not imply that there is no overlap; on the contrary, it is often the case that the maker of such claims offers him- or herself as the representative of a certain constituency, which can be the same as the intended audience. In this context, Saward contends that political representatives pick their constituents ‘just as much as’ the electorate chooses their representatives—by referring to some groups rather than others, depicting them as ‘this or that, as requiring this or that, as having, this or that set of interests’ (ibid, p. 301).

As a variant of media content analysis which connects ‘makers’ to policy positions, claims analysis is the most appropriate method for this type of investigation because it allows the identification of all elements in Saward’s framework, including the audience (de Wilde 2013). Thus, Saward’s idea of ‘maker’ is understood as the ‘claimant’ in Koopmans and Statham’s model, i.e. the person/entity doing the talking in the public sphere. The ‘subject’ preserves its meaning as the ‘representative’ of a given constituency, presented as such by the claimant. The ‘object’ is the intended ‘constituency’ in the name of whom the representative claim is being made, while the ‘audience’ are the ‘addresses’ who not only witness but in some cases are expected to act upon the claim (ibid., p. 286).

The following section introduces the main parameters of the case study on which this method is applied and then discussed.

Beyond the Euro crisis: the politicization of free movement of people

On Sunday 17 April 2011, in the small Italian city of Ventimiglia, close to the border with France, the European free movement regime entered a crisis. In the context of the Arab Spring and the ensuing migratory pressures in the Mediterranean, an influx of Tunisian refugees reached the shores of Italy (Pascouau 2011). Overwhelmed by the inflow and denied help from other member states—despite repeated solidarity pleas—Italian authorities decided to grant residence permits to those refugees seeking to cross the border to France (Paulic 2011). Since French and other national authorities did not recognize the legality of the documents, human rights activists got involved in the dispute, offering to accompany the Tunisians to France in a so-called Dignity Train that ran from Genoa to Marseille via Ventimiglia. Claiming that their demonstration was unauthorized and posed threats to ‘public order’, the French administrative authorities decided to stop the trains in Ventimiglia and conduct temporary border controls in order to return the unwanted immigrants. The incident immediately escalated at the political level, as French President Nicolas Sarkozy and Italian Prime Minister Silvio Berlusconi met a few days later in Rome and declared that the cause of the bilateral dispute was the absence of lawful mechanisms to reintroduce internal border controls in the Schengen Area. The European Council addressed this crisis during its regular June meeting and endorsed the necessity of a reform to the Schengen Borders Code, setting the deadline of September for the introduction of a Commission proposal (Berthelet 2012, pp. 658–661).

The Commission complied with the deadline, but produced a surprise by interpreting the European Council conclusions to its own advantage and shifting the responsibility for the reintroduction of border controls from member states to itself, which infuriated national representatives in the Council (Changeur 2011b). To enact this clever move, the Commission selected a specific legal basis which additionally involved the Parliament in full capacity as co-legislator; consequently, when the Council unilaterally decided to change the legal basis in June 2012, they effectively excluded the Parliament from the decision-making process, which in turn infuriated the Members of the European Parliament (MEPs) (Pascouau 2012). In retaliation, the latter decided to freeze negotiations with the Danish presidency of 2012 on five legislative files considered important by the Council, in the hope that such ‘show of force’ would make ministers reconsider their decision about the legal basis (Vandystadt 2012a). After a few months of uncertainty, the Council proposed the Parliament to continue negotiations as if they were debating a co-decision file in which the rapporteur’s opinions were duly taken into consideration, but without changing the legal basis. Once trilogues started, discussions advanced quickly and the file was concluded in May 2013, with formal adoption in October of the same year (Peers 2013, p. 30).

From beginning to end, the legislative file—formally known as the Schengen Governance Package—was deeply politicized, with a strong media focus on the Franco-Italian crisis which later moved to a lesser extent to the inter-institutional conflict. Indeed, the media spotlight on Schengen is not surprising given the symbolism and tangible benefits of free movement in the eyes of ordinary EU citizens. Against this background, the initiative from the spring of 2011 to allow the reintroduction of internal border controls stirred controversy across both Schengen member states and EU institutions. Rolling back the Schengen system in this way was not only a matter of administrative border management, but raised sensitive questions regarding what Schengen is and how it should consequently be protected or enabled, which is why it became politicized. This was evident from the types of claims put forth in the media, which are discussed below.

Drawing the map of representative claims: how free movement became politicized

Taking into consideration the unfolding of the case—which shifted from being a bilateral dispute to an EU-level inter-institutional conflict—the empirical material examined is limited to publications covering the EU; this choice is based on the observation that national media reported on the Franco-Italian affair but mostly overlooked the ensuing EU inter-institutional conflict. Two publications—Agence Europe and Europolitics—are sampled for case-related articles; both considered authoritative and independent sources of news about the EU.Footnote 2 A total of 96 articles were analysed over a three-year period (2011–2013), out of which 28 did not include any representative claims. (They were just describing facts.) Conversely, 73 representative claims were identified, the number being so high because some articles included two representative claims (from different institutional actors). If a representative claim came from several claimants within the same institution, only one code was assigned. The highest number of articles (59) was found in 2011 in the context of the Franco-Italian affair and the subsequent European Council meeting and European Commission proposals on the topic. Fewer articles (33) were published in 2012 following the legal basis debacle and the ensuing inter-institutional conflict, while in 2013—when the file was concluded—only four articles were found.

Figure 1 offers an overview of the main elements of the representative claims framework applied to the Schengen Governance Package. Following Saward’s model, I identified the makers (putting forward the representative claim—the ‘claimants’); the subjects (the institutional actors doing the representation, a.k.a. the ‘representatives’); the objects of representation (i.e. the constituencies claimed to be represented); and the audiences (the explicit/implicit addressee of the claim). Five types of claimants have been detected, most often speaking on behalf of their respective institutions: (a) MEPs on 34 occasions; (b) members of the College of Commissioners (i.e. the Commission President or the Home Affairs Commissioner) on 13 occasions; (c) Interior Ministers from national governments on 15 occasions; (d) Presidents or Prime Ministers of member states on 10 occasions; and (e) the President of the European Council (on 1 occasion). Furthermore, the representatives are identified as the institutional actors under consideration: (a) the European Parliament; (b) the European Commission; (c) the Council, (d) and the European Council. Very importantly, within the intergovernmental forums, we often see representatives of national governments (ministers or Prime Ministers/Presidents) speaking in the name of their respective governments—e.g. Sarkozy representing the French government—which is why in that circumstance they cannot obviously be considered to represent an EU institution.

Fig. 1: Schengen Governance Package—Map of Representative Claims
figure 1

Own account, based on Atlas.ti network analysis of 96 articles from Agence Europe and Europolitics related to the Schengen Governance Package during 2011–2013. The number between brackets illustrates the number of quotations associated with the code

Full size image

Figure 1 shows how constituencies have been empirically derived into: (a) voters sympathetic to asylum/immigration (mentioned 10 times, mostly by the Parliament); (b) free movers in Schengen (invoked 47 times by the Parliament, the Commission, and the European Council); (c) asylum-seekers (defended explicitly by the Commission once); (d) national law enforcement/border authorities (whose interests were cited 5 times by national representatives); (e) national governments (mentioned on 5 occasions at the level of ministers and Prime Ministers/Presidents); and (f) national voters opposing immigration (whose preferences were translated on 15 instances by representatives of their national governments). Last but not least, the audiences have proven very diverse and can overlap, ranging from: (a) EU citizens, to (b) EU-level non-state actors (non-governmental organizations, academics, or think tanks engaged in the policy debate)—exclusively addressed by the Parliament and the Commission; moving to (c) other EU institutions addressed by each representative in turn (e.g. the Parliament signals its position to the Council or the Commission); going to (d) national governments addressed most often by their own members (e.g. an Interior Minister signals his/her position to other members of the Council); and finally arriving to (e) national citizens targeted—predictably—by their national representatives (e.g. Sarkozy speaking about Schengen to French citizens during the presidential campaign).

Having outlined the main elements of the conflict, the discussion now moves to the mechanisms of politicization, analysed according to Statham and Trenz’s framework.

Polarization and legitimation

At the surface, this case illustrates three subsequent conflicts: (1) between member states in the European Council and the Council, particularly Italy and France, whose leaders and relevant authorities were throwing the burden of refugees from one to another; (2) between the Commission and the Council, who had different visions of whom should decide on the temporary reintroduction of border controls, and (3) between the Council and the Parliament, who disagreed on the decision-making mode through which such legislation should be enacted. At the substantive level, the way in which the Schengen Governance Package was politicized reveals important differences of opinion between the various actors involved regarding the very function of the Schengen Area, as captured in the following quotes:

Schengen exists to defend freedom of movement. It is not an instrument to control migratory flows. (Cecilia Malmström, Home Affairs Commissioner 2012).

Schengen is not just a treaty on the free movement of persons, it is also a treaty on the protection of external borders. (Claude Guéant, French Minister of Interior 2012).Footnote 3

Freedom of movement across EU countries is not a national right. It is a fundamental right guaranteed to EU citizens by the Treaties, which prevails over temporary national interests. Accordingly, no national government alone can take decisions affecting that right without them first being agreed at EU level. (Changeur 2011a).

Indeed, it appears that a fundamental cleavage crystallized here, but the type of polarization went beyond pro- and anti-integrationist lines. At first glance, one could say that the request to regulate the reintroduction of border controls at the EU level was in itself a form of rolling back supranational integration through ‘a complete reversal of the Schengen philosophy’ of mutual trust and solidarity (Pascouau 2011, p. 2). In this respect, the position of some member states in the European Council and the Council, led by France, could be construed as anti-integrationist. It is important to mention, however, that in this case a European issue (the absence of internal border controls) became caught in a broader pro-/anti-immigration debate at the national level, meaning that the crisis was not about the European issue to begin with. Following the Ventimiglia incident, France demanded a reform of EU rules that would grant national authorities more control over anti-immigration efforts, while Italy agreed to the request in the hope to obtain (financial) solidarity in the management of refugee inflows; other member states—including Germany, the Netherlands, Belgium, or Austria—expressed concern at the Italian decision to grant residence visas to Tunisian refugees, but did not move to reintroducing border controls (Carrera et al. 2011, p. 2).

The point is that some member states were clearly taking a nationalist stand in respect of how immigration crises should be handled—by seeking to bring back national borders or dislocate the burden of refugees to their neighbours—and this had important EU-level consequences. The sort of conflict that followed resembles very closely the demarcation-integration cleavage born out of cultural diversity in contemporary globalized societies, as identified in the literature on political conflict in Western Europe (Grande and Kriesi 2015, pp. 193–5). Specifically, the reason why some political actors—most vocally French President Nikolas Sarkozy and his Interior Minister Claude Guéant—supported the reintroduction of border controls was because they aimed to gain political capital among right-wing anti-immigration voters in France, even if this meant bringing a European issue (the free movement of people) into to the political debate. For example, before the European Council meeting in June 2011, Sarkozy was only speaking of the necessity to make changes to the existing Schengen system (Spiegel Online International 2011), while in the later days of his 2012 presidential campaign he declared that he was ready to withdraw France from the Schengen system altogether if the EU does not strengthen its external borders (Paulic 2012).

In response to the strong position expressed by the French, other national governments had to position themselves. At the June 2011 meeting of the European Council, Polish Prime Minister Donald Tusk declared that debates between member states were ‘fierce’, with some advocating the priority to combat illegal immigration by allowing internal border controls, while others—Poland included—defending the right of EU citizens to free movement (Kordys 2011). Accordingly, the conclusions called on the Commission to propose a legislative package by September that ‘would allow the exceptional reintroduction of internal border controls in a truly critical situation’, ‘without jeopardizing the principle of free movement of persons’ (European Council 2011). The solution proposed by the European Council was a compromise trying to appease member states on both sides of the conflict and allow heads of state and governments to focus on more pressing issues related to the economic crisis.

Nevertheless, from the moment the issue reached the agenda of supranational institutions—the Commission and the Parliament—their position was unfaltering. For MEPs, Schengen was evidently linked to the free movement of EU citizens, which is why they saw the whole proposal as unfounded (European Parliament 2011). For its part, the Commission chose in a first step not to antagonize member states and did not open infringements procedures against either Italy or France for the Ventimiglia incident, although there would have been legal basis for such an action (Carrera et al. 2011). The Home Affairs Directorate-General became much more articulate on the matter starting September, when the Commission submitted a legislative proposal as instructed by the European Council but altered its substance by transferring decision-making authority on the matter from the Council to itself. As expressed by Home Affairs Commissioner Cecilia Malmström:

Free movement is in the common interest and it is felt that the decision to reintroduce border controls should therefore be taken at European level. The Commission proposes a mechanism for a coordinated EU response to protect the functioning and the integrity of the Schengen Area. (Changeur 2011b).

The trick was in the legal basis, selected by the Commission as article 77(e) TFEU, related to ‘the absence of any controls on persons, whatever their nationality, when crossing internal borders’. In other words, the Commission proposed the reintroduction of border control as an aspect of the free movement acquis to be decided upon by the Council and Parliament in line with the ordinary legislative procedure (Pascouau 2012). The Council was far from agreeing with this assessment, advocating article 70 TFEU as the appropriate legal basis, according to which the Council was responsible for ‘laying down the arrangements whereby member states, in collaboration with the Commission, conduct objective and impartial evaluation of the implementation of the Union policies (…) by member states’ authorities’. The legal basis problem was as much administrative as it was political, as member states considered that the Commission was attempting to infringe upon their national sovereignty by seeking ‘to assume responsibility for making decisions on operational measures in the security field’ (Changeur 2012).

Subsequently, another line of conflict became apparent in the crisis, which was not so much about pro- or anti-European camps as it was about what kind of ‘Europe’ is desirable: a supranational project where the free movement of persons is guaranteed to all EU citizens (a position held by the Commission and the Parliament), or an intergovernmental project in which member states are responsible for ensuring the security of their national citizens (a position shared by most member states in the Council). While representative claims were made on both sides, the extent to which this polarization resonated with the public remained limited, as discussed in the next section.

Public resonance and legitimacy

National media paid close attention to the Franco-Italian dispute of 2011: the issue featured strongly in mainstream French newspapers (Le Monde, Figaro, Le Nouvel Observateur) and Italian ones (La Repubblica) but also German (Süddeutsche Zeitung, Der Spiegel) and English publications or news outlets (The Telegraph, the BBC) (Lerougetel 2011; Spiegel Online International 2011). The two EU-level publications analysed here—Agence Europe and Europolitics—also covered the diplomatic row between France and Italy the most (59 out of 96 articles). As shown in Fig. 1, the media captured well the polarization between the ‘anti-immigration/reform-Schengen’ camp (composed of some member states in the European Council and the JHA Council, particularly Sarkozy and associates) and the ‘pro-refugees/preserve-Schengen’ camp (Poland most vocally, the President of the European Council, the Commission, the Parliament). A parallel conflict on decision-making procedures and competences emerged later between what could be called an ‘intergovernmental Schengen’ camp (made of the Council) and a ‘supranational Schengen’ camp (made of the Commission and the Parliament). The two were overlapping only to a certain extent, as not all member states in the Council were in favour of bringing Schengen in the anti-immigration rhetoric, albeit a majority wanted to maintain the system as intergovernmental as possible.

It is difficult to say, however, that media attention to the crisis has been a factor in its own right. It seems more likely that the French President and his Interior Minister—who started the claim-making process in the media to begin with—sought to address their own national audiences in view of the 2012 presidential campaign; for example, their most radical claims were put forth in the context of electoral rallies. The Schengen system of rules was the collateral damage of the electoral campaign, and the extent to which the discussion would have become so polarized in the absence of Sarkozy’s bid for re-election is questionable. For example, a wide majority of the representative claims made by the Commission, MEPs, and the President of the European Council were in reaction to Sarkozy’s direct threats to the free movement regime. On a different note, Berlusconi and his Foreign Affairs Minister Franco Frattini, also often cited in the media, seemed keen to emphasize the lack of solidarity from other member states in the face of the immigration crisis hitting Italy—in an obvious attempt to address the concerns of their own electorate to the migration inflows. To put it differently, the media seems to have been instrumentalized for electoral purposes and ended up transposing or reacting to ‘populist pressure’ (Spiegel Online International 2011).

In this context, the main actors of the crisis were politically elected figures (part of national governments or the European Parliament), together with Commission and European Council representatives. The ‘public’—composed of citizens, civil society organizations, or social movements more broadly—was the great absentee from the media. We only know what citizens wanted from how they were portrayed in the representative claims of various officials: whether they were pro- or against immigration, whether they thought free movement of people should be protected, etc. In the end, Sarkozy lost the election in France despite his consistent efforts to appeal to anti-immigration voters of the Front National. In addition, the fact that the second phase of this case, including the EU inter-institutional disputes, was much less reported in newspapers is indicative of the limited role of the media as an independent actor. While the Franco-Italian affair was attractive owing to its ‘scandalous’ nature, the inter-institutional debate about which type of Schengen regime should be preserved was far less sensational owing to its dry elements, namely the discussion over the legal basis. In other words, the role of the media as the ‘transmission belt’ from the public to decision-making actors failed to function in this case.

Conclusion

This article investigated an earlier case of an EU immigration crisis in order to evaluate the applicability of democratic functionalism as a theory explaining mechanisms of EU politicization. First, it found that that the 2011 Franco-Italian crisis over the management of Tunisian refugees had indeed opened up discursive opportunities about the nature of the border-free regime in the EU, but that polarization did not occur along pro/anti-EU lines as expected by democratic functionalism. Instead, by relating the Schengen reform to the fight against illegal immigration, French President Nicolas Sarkozy tried to appeal to those ‘individuals who possess a strong identification with their national community and who are attached to its exclusionary norms [and thus] will perceive a weakening of national institutions as a loss’ (Kriesi and Grande 2016, pp. 193–4). Furthermore, while the media definitely played a crucial role as the medium through which different actors put forth competitive claims about the issue at stake, it cannot be considered an ‘independent, self-steering’ (Statham and Trenz 2015, p. 290) entity in the case because of the way in which it was obviously instrumentalized by the Sarkozy camp in a (pre-)electoral effort connected with his 2012 bid for re-election. Other governments and EU institutions reacted to his ‘tough’ stance on immigration, which would have affected free movement of people in the Schengen Area. Thus, the debate shifted from the possible reintroduction of frontier checks to the very purpose of the Schengen space and how it should consequently be protected or enabled. Finally, the ‘public’ was conspicuously absent from the debate, as the feedback loop from citizens to political élites failed to materialize. One could only observe how political representatives portrayed the interests of their constituents as being for or against immigration and the abolition/maintenance of free movement of people. This marks a significant difference from the Euro crisis, which has seen a broad political mobilization of non-governmental and non-EU actors.

To what extent does this argument apply in relation to the 2015 refugee crisis? First, unlike the 2011 Franco-Italian dispute, the 2015 episode involved both a higher number of refugees arriving to the EU and more member states affected. The politicization that followed took different dimensions depending on whether member states were on the frontline of the crisis (Greece, Italy), whether they were destination countries (Germany, Sweden), whether they were transit countries (Hungary, Slovenia), or whether they were bystander countries not directly affected because of their off-route location (Schimmelfennig 2017, p. 8). EU institutions and governments of frontline and destination countries attempted to place Schengen solidarity at the centre of the debate; however, economic and social issues were rapidly hijacked by ‘identity’ politics and the question of who ‘belongs’ to the European society (Börzel and Risse 2017, pp. 18–19). A European nationalist, anti-immigrant, and anti-Islam stance was particularly prominent in the discourse of right-wing populist parties such as the French Front National, the German Alternative für Deutschland, the Polish Law and Justice Party, the Hungarian Fidesz, the Swedish Democrats, the Danish People’s Party, the Austrian Freedom Party or the Dutch Party for Freedom (The Guardian 2016). This finding corroborates the conclusion of this article regarding the politicization of EU immigration crises along a cultural integration-demarcation cleavage (Kriesi and Grande 2016) rather than along pro-/anti-EU lines.

What differed in the 2015 episode from the one investigated in this article was the instrumentalization of the media and the mobilization of the public. Given the magnitude of the 2015 refugee inflow, the crisis featured heavily in mainstream newspaper coverage across member states (Georgiou and Zaborowski 2017), while immigration became a top concern for EU citizens according to the 2016 Eurobarometer survey (Debomy and Tripier 2017). This important difference from the 2011 crisis suggests that democratic functionalism might be applicable after all—with some revisions. To correct its flaws, the theory needs to specify (1) how big a crisis or how much EU politicization is necessary to generate sufficient media attention and mobilization from the public, and (2) for how long a European issue must be contested in the public sphere to display the ‘mechanisms of politicization’ expected by democratic functionalism. In other words, if the politicization of EU immigration crises requires time to build up, democratic functionalism needs to add a temporal dimension to its main tenets.