1 Introduction: Do External Control Policies “Converge”?

The capacity to control geographical mobility across political borders is a key aspect of sovereignty among modern states. Unauthorised movements across borders are consequently seen as a challenge to their very raison d’ être. Migration has always raised a variety of existential fears: the control of the intra-European labour supply had triggered bellicose concerns already in the period leading up to World War I (Olsson, 1996). The fear of unmanageable “surplus populations” has accompanied the extensive redrawing of Europe’s maps and all the unmixing of its empires (Gatrell, 2019). More recently, the (allegedly) inadequate control of European borders has been described as a clear and present danger, a threat to the survival of the European project and even European civilisation itself.

These fears may appear far-fetched, but public opinion reveals a different story. Opinion polls across Western European states have consistently shown the existence of a sizeable, and increasingly easy to mobilise, bloc of voters opposed to further immigration. The deep restructuring of European party systems in the aftermath of the 2015–16 asylum wave provides further evidence for the intrinsic appeal of playing the “immigration card” for a variety of populist challengers.

Because the effectiveness of border controls is seen as an important attribute of sovereignty, the claim that some states are better than others at securing their borders takes on a strongly normative connotation. Global migration control is depicted by some scholars as very uneven, a world in which some states are highly effective in policing their borders, while others are unable to control unauthorised migration (if not even colluding with it). The spectre of a radical difference between Northern and Southern states (with the former “good” at managing migration, and the latter unable to provide effective control) has haunted Western Europe since the very beginnings of a distinctly “European” migration policy. Migration control in Western European countries is in fact highly interlinked, and the enduring presence of unwanted migration is often considered the consequence of ineffective action – or even inaction – by the (allegedly) “inexperienced” Southern partners.

In the aftermath of the migration “crisis” of 1989–91, Southern European countries were openly described as both inefficient and lacking the necessary experience in terms of mechanisms of immigration control and humanitarian protection (Baldwin-Edwards, 1999, 2001; Thränhardt, 2003). They were associated with lax migration and border controls, weak or non-existing asylum programs, and a high degree of public ambiguity towards irregular migration. Northern European countries, in contrast, were considered to have reliable and efficient mechanisms of migration control and refugee protection, as well as little tolerance for irregular immigration (ibid). The popularity of such a vision not only derived from widespread stereotypes, but was also politically important, giving priority to the interests of Northern European countries. It cast the countries trying to enter into the Schengen Agreement in the role of unruly pupils that needed to be disciplined (Baldwin-Edwards, 1999, 2001).

In 2001, Baldwin-Edwards described the Mediterranean countries as the “weak underbelly of the EU control system” (2001, p. 23). Southern countries were basically transit countries used by the masses of unwanted migrants to gain a foothold in Europe. Passing through these countries, migrants could prepare to move towards the “real” migration destinations in the North (Baldwin-Edwards, 2001). Southern European migration systems, which were gaining momentum precisely in those years, were thus interpreted as “trouble” for Northern European countries, rather than as independent migration systems in their own right (Sciortino, 2005).

The idea of a North-South divide in the effectiveness of migration control has been, however, contested on several grounds. By comparing Italy and Germany (typically considered the two showpieces of “soft” and “hard” lines in immigration control policies), Finotelli (2009) concluded it was a myth. She documented how the existing differences both in refugee reception and irregular migration flows could be better explained as the outcomes of different inclusion and reception mechanisms, rather than by inefficient and lax border controls (Finotelli, 2009). Furthermore, as Finotelli and Sciortino (2009) have argued, the functioning of mechanisms for immigration control must be assessed within the historical development of a given country’s migration regimes. From a different angle, it has been shown that Mediterranean control systems have been reasonably effective (Colombo, 2012).

The idea that EU Member States may be distinguished according to their control effectiveness is not limited to the idea of a cleavage between Northern and Southern states. It also provides the background for the ever-popular debate concerning the existence of convergence (or lack thereof) among “core” and “outer” members (Meyers, 2002; Toshkov & de Haan, 2013; Hollifield et al., 2014). In several migration policy fields – labour migration, integration policies and asylum, in particular – many scholars have sought to identify a trend pointing (or failing to point) towards increasingly similar policies and outcomes. Scholars have highlighted how, beyond country-specific variations, there exists a set of mechanisms – operating across Western European states – that account for increasingly similar policy outcomes (e.g., Meyers, 2002). EU Member States, (indeed, all industrially advanced states) converge on a very similar approach: “Courting the Top, Fending-off the Bottom” (Joppke, 2021, p. 68).

With regard to national asylum policies, and, specifically, overall asylum recognition rates by European states, a study by Toshkov and de Haan (2013) supports the convergence hypothesis. In their view, such increasing similarity is linked to the creation and consolidation of the Common European Asylum System (CEAS).Footnote 1 Even as regards hyped and heated integration debates, some scholars have pointed to a slow convergence towards similar patterns of “civic” integration inspired by repressive liberalism (Joppke, 1998).

A main stumbling block for debates on convergence, however, has been the lack of evidence on increased systematic similarities among the EU Member States. The above-mentioned claims, in fact, have been quickly challenged by contrasting studies. Analysing five European countries from 1990 to 2016, Consterdine and Hampshire (2020) find scarce evidence of a general change in the direction of restrictive (or liberal) labour migration policies. Some scholars have identified different degrees of integration policy change at the national level, making the dividing line among national regimes more blurred than it appears at first glance (Finotelli & Michalowski, 2012; Caponio et al., 2016).

In this chapter, we analyse the similarities and differences among EU Member States in two migration policy fields that have not received much analytical attention: visa policy and deportation/return policy. We have chosen to focus on policy fields widely different in terms of history and policy development. Visa policy is likely the oldest and most stringent area of coordination among EU Member States. Deportation/return policy, in contrast, is an area in which, besides ritualised statements, supra-national interventions have been flimsy, if not utterly contradictory. As both visa and deportation/return policies play an important role in the European migration control system, the institutional differences among them are compelling.

2 A Critique of the Implicit Conceptual Framing of Debates on Convergence

This chapter provides an empirical critique of the implicit conceptual frame of increasingly polarised debates on convergence among EU states. This lack of agreement and reasonable dialogue leads to a scenario in which important differences among EU Member States are overlooked. Equally important, the belief in the existence of deep differences among Member States plays an important role in European migration policy making. It supports a very simplified, and unrealistic, interpretation of the social dynamics of unwanted migration, making more likely the chances of conflict and lowering the chances of adopting adequate solutions.

The debilitating weakness of these debates is the (implicit) assumption that the migration control strategies of EU Member States could, at least in principle, be fully independent. The development of an ‘EU migration policy’ is consequently measured out against an (unspoken) ideal in which states renounce such independence to adopt a standardised and uniform “EU” policy. As a category of practice (Bourdieu, 1997), this assumption is clearly significant. National politicians are always ready to claim the existence of “national” control goals thwarted by “European” rigidities. “European” politicians and bureaucrats, similarly, are always happy to point to the “egoism” of states as impeding the rational development of an adequate migration control system.

Categories of practice, however, hardly ever work satisfactorily as category of analysis (Bourdieu, 1997). In fact, there are several reasons for defining the European migration regime not as a set of Member States with a super-imposed authority, but rather as a highly interdependent and yet politically highly segmented system (Bommes, 2012). The interdependence among European powers with regard to migration policy is much older than any EU attempt to “regulate” the phenomenon. It is much older than the Union itself. Already during the Huguenot crisis in the 1680s, European kings and princes had tailored their admission choices through careful anticipation and monitoring of what other powers would do (Orchard, 2014). More recently, the so-called Tamil “crisis” of the early 1980s – when asylum seekers were arriving from Sri Lanka to the German Federal Republic and France (and ultimately to Canada) through East Berlin – represents an especially important lesson for the design of any subsequent migration policy (Sciortino, 2017). No European state has ever been able to control its borders autarchically.

In the end, it is simply impossible to analyse migration control policies and their outcomes for each European state individually. Because they are part of a system, the differences and similarities among them can be understood only by looking at the role each state plays in it. If the numbers at the border between Belgium and the Netherlands are low, this is not because Dutch migration controls are somewhat more “effective”. It is simply because many other Member States are willing – enthusiastically or not – to apply similar visa requirements to the citizens of some sending countries, preventing transit migration. If some Mediterranean countries experience strong pressures over their maritime borders, this is not because their border controls are “inefficient”. In fact, boats are identified long before their landing. Boats arrive on Mediterranean shores because the entire EU control system is designed to make some Mediterranean corridors the only available option for those entering Europe to claim asylum. Moreover, international legal protection for refugees makes it impossible for Mediterranean states to push them back once they have arrived on national territory. This scenario stems from the fateful decision, at the Tampere meeting, of binding European migration policy to “full and inclusive application of the Geneva Convention thus ensuring that nobody is sent back to persecution, i.e., maintaining the principle of non-refoulement the legal protection guaranteed by the non-refoulement principle of international refugee law” (Presidency Conclusions, 1999).

To analyse the development of European migration policy – and the strains it reveals – it is necessary to abandon the rhetoric of “effective” vs. “weak” states, to focus instead on functional and segmentary variations. The importance of such a distinction becomes clear when we acknowledge and confront a field in which European states have achieved a large degree of supranational integration, i.e., in terms of visa policy (Finotelli & Sciortino, 2013; Nicolosi, 2020), with a field in which formal harmonisation and cooperation among European states remain very low (return and removal policies).

In both cases, however, careful analysis reveals a similar picture: the differences between Northern and Southern countries are not particularly strong or clear-cut. Most of these differences may be accounted for by functional variations, by the different roles states play within the European system of migration control. As far as visas are concerned, we find a slow process of inclusion for Southern (and Eastern) countries in a control mechanism long shaped by the overall control objectives of Northern European countries. In contrast, return and removal policies have been – and largely are – intentionally kept outside of European coordination (De Bruycker et al., 2016). Even in this case, however (and even if the data available are rather spotty), there is no systematic evidence of a North-South dichotomy. If there is a similarity to be detected, is the generalised low level of effectiveness in removing unauthorised third-country nationals (TCNs) across all European states.

3 A Tale of Policy Convergence: Short-Term Visas as a Generalised System of Migration Control

Many of the tools used today by states for controlling mobility are little more than a century old. Consider the case of the travel visa, which had come to be used as a generalised system of migration control only at the beginning of World War I (Czaika et al., 2018). Its salience in European migration history has shifted considerably along the different periods of European migration history.

In the two decades after World War II, Western European migration policies were confronted by the presence of a “surplus population” comprised of approximately 11 million internally displaced persons (IDPs), refugees and asylum seekers. This was further augmented by a “reflux” of settlers from Eastern Europe, the Balkans and newly decolonised countries (Peach, 1997). Many European citizens (including many national leaders) had direct experience with exile and forced displacement. The pressures of the Cold War made refugees living proof of the superiority of the “free world”.

All these elements contributed to the introduction of a highly liberal regime, anchored in explicit provisions in new national constitutions and in the adoption, in July 1951, of the Convention Relating to the Status of Refugees (Gatrell, 2000; Rinauro, 2009). The Convention provided, although initially only for European refugees after World War II, a clear definition of refugee, sharply differentiating them from migrant workers. It further established the principle of individual protection and the binding obligation of non-refoulement. The main priority for European migration policy at that time was securing visas to allow as many refugees as possible to leave Europe. Categorised as a “surplus population”, their extra-European mobility was seen as essential for the stabilisation of the European continent. Very little attention was paid to regulate new arrivals on the continent, as they were considered rare. With the consolidation of the communist bloc, flows from Eastern Europe were severely curtailed, making the fear of new arrivals in Western Europe quite limited.

The situation changed with the European “economic miracle” of the 1950s and 1960s. In a context of extraordinary economic growth, securing an adequate supply of foreign labour in the form of a low-skilled workforce become a pressing concern (Judt, 2006; Bernard, 2019). This “influx” phase – to use Peach’s (1997) periodisation – was characterised, primarily, but not exclusively in Northern European countries, by the very selective use of visa requirements – largely absent, or informally ignored in the case of citizens from certain origin countries, rigorously enforced, even beyond diplomatic agreements, against the citizens of others (Schönwälder, 2001).

This phase ended with the oil shock of the early 1970s, accompanied in Western Europe by the interruption of all active programs for low-skilled labour recruitment (Bernard, 2019). With the adoption of an increasingly restrictive approach, several countries experimented with the use of visa requirements as a tool to prevent unwanted migration. In the 1980s, visas become a central element in migration control, targeting flows of potential asylum-seekers. Two factors converged to make Western European public opinion increasingly adverse to refugees. The first was the sharp increase in the number of asylum seekers. From 1970 to 1999, the number of asylum applications in Western Europe increased dramatically, from 15,000 to 300,000 a year (Hatton, 2004; Van Mol & de Valk, 2016). The second was the diversification among places of origin, with a growing number of non-European asylum applicants. They were often blamed of not being “real” refugees but rather migrant workers in disguise, trying to compete unfairly with natives in the national labour markets or abusing national welfare systems (Van Mol & de Valk, 2016; Sciortino, 2017).

Given the protection granted by the international refugee regime (especially by the non-refoulment clause), the prevention of unwanted flows of asylum-seekers requires barring them from arriving on the territory of a state in which they could claim protection. The introduction of visa requirement was an especially convenient control tool. In the Western European context, Germany was the first country to experiment systematically with visa requirements for the prevention of unwanted migration flows of Bangladeshis, Indians, and Sri Lankans in 1980 and Ethiopians in 1982. A few years later, France and the United Kingdom followed suit. Such decisions produced, in the short term, the required effect – the number of asylum-seekers decreased (UNHCR, 2011).

Very quickly, however, the limits of country-based visa policy became evident. The emblematic case occurred in Germany between 1980 and 1985. When the Federal Republic of Germany (FRG) introduced visa requirements for Bangladeshis, Indians, and Sri Lankans, the DDR began to encourage potential asylum seekers to fly visa-free to East Berlin, from where the potential applicants could easily reach the FRG (since the city was still considered a single administrative unit by West German laws). The FRG was able to contain the arrivals from those areas only when, in exchange for sizeable amount in loans, the DDR agreed to introduce similar restrictive actions. In short, since its very beginning, the effectiveness of visa policy for preventing the arrival of asylum-seekers was strictly contingent upon the willingness of neighbors to participate in the action.

Such actions were at the centre of intergovernmental cooperation in the field of Justice and Home Affairs, starting with the Schengen Agreement in 1985 and the following Convention in 1990. The original signatory states – only Belgium, France, Germany, Luxembourg and the Netherlands – were, unsurprisingly, the largest receivers of asylum-seekers in Europe (Hatton, 2004; Van Mol & de Valk, 2016). The Schengen Agreement, considered nowadays a key milestone in establishing an internal market with the free movement of persons, was an objective consolidated by the Treaty of Maastricht in 1992. These actions have made possible the abolition of internal border controls for all nationals of Europe’s Schengen Area, today, all EU nationals. For our purposes, however, the Schengen Agreement represents, above all, the establishment of a system of external border control, common to all states adhering (and desiring to adhere) to the Schengen Area. Members and prospective applicants have been required to introduce visa requirements for the citizens of several non-European states (TCNs) and to accept a collective procedure for selecting those that could enjoy visa-free travel. In addition, states who wished to adhere to the Schengen Agreement were required to introduce sanctions for all the carriers transporting irregular migrants, adopting ad hoc asylum policies, strengthening border controls, developing more severe measures against irregular migrants in the national territory, and contributing to a common dataset of all detected irregular migrants (Sciortino, 2017).

This set of measures was a reaction to an important bifurcation existing at the time among Western European states. Because visas were strongly associated with the prevention of asylum-seekers, Northern European states, at the time, the nearly exclusively targets of asylum applications, were imposing many visa requirements. Conversely, Southern European countries, where the number of asylum-seekers was negligible, maintained many visa-free agreements (Finotelli & Sciortino, 2013). Most of the migratory flows to these countries consisted of seasonal migrant workers who played an important role in supporting national economies, especially with agricultural labour, and care work in households and elderly people (Cvajner, 2012; Sciortino, 2017).

From the point of view of the original Northern signatories, the highly discretional inclusion of Southern states in the Schengen system was to be balanced by their willingness to participate in the control strategies of Northern European countries. One of the most feared side effects of the pact was the possibility of potential asylum seekers flying or landing in Southern countries and subsequently moving more freely across all Schengen states, especially those in the North. The German delegation was particularly explicit about these concerns, asking for an additional annex to the Agreement, in which the abolition of internal border controls was conditioned to the introduction of visa requirements for TCNs and other compensatory measures. These measures were presented as necessary for safeguarding internal security and the strengthening of European cooperation against unauthorised migration and asylum flows (Finotelli & Sciortino, 2013; Paoli, 2018).

The objective of integrating neighbouring countries into the prevention of the arrival of asylum-seekers was complemented by the Dublin Convention (1990, then modified in 2003 and 2013). The Dublin Convention, meant to prevent what was dubbed “asylum-shopping”, was designed to make countries with external borders responsible for the management of asylum-seekers who, unable to receive a visa, would try to reach the territory of the EU.

The Schengen Agreement and the Dublin Convention provide clear evidence that the development of the European visa regime may be considered an outward diffusion process governed by the migratory interests and control goals of core Northern European countries. Southern European states – and much later, Eastern states participating in the Eastern Enlargement – have progressively shaped their control policies and practices in accordance with these goals.

This process did not occur without resistance. Indeed, Southern countries have been often reluctant to accept the Northern model, especially because the flows of irregular migrant workers to these countries have played an important role in supporting national economies and welfare (Sciortino, 2017). Nevertheless, being an inter-governmental initiative, insider Northern states were able to exert pressure on Southern candidates, pushing them to introduce stricter visa restrictions as a precondition for participation in the Schengen process (Sciortino & Finotelli, 2013; FitzGerald, 2019). The Southern expansion of the Schengen system started with the participation of Italy in 1990, Portugal and Spain in 1991, and Greece in 1992 (Paoli, 2018). Joining the Schengen “club” required introducing visa requirements for countries that had a long history of unencumbered travel, such as the countries of the Southern rim of the Mediterranean (for instance, the case of Italy for citizens from Tunisia and Turkey) and those with whom they had historical and colonial ties (the case of Spain and many Southern American states).

In 1992, the Treaty of Maastricht reaffirmed the key role of intragovernmental cooperation among European states on migration policy, locating it within the third pillar (Justice and Home Affairs) of the Treaty on European Union (TEU). In 1995, when seven member countries established effective border-free travel among them, Regulation No. 2317/1995 of the European Council introduced a first common list of 101 countries whose nationals were required to obtain a visa to enter the EU. The Amsterdam Treaty, in 1995, provided, in one of its protocols, for the transfer of the Schengen acquis into the legal and institutional framework of the EU (Peers, 2000). In simple terms, the process of visa harmonisation was further strengthened by taking on a supranational shape. Namely, states and perspective members are now formally bound to implement Schengen rules as part of the pre-existing body of EU law that any applicant is obliged to accept.

A further fundamental step took place in 2009, with the adoption of the Visa Code by the Regulation of the European Council No. 810/2009. The Code systematised the visa application procedure, by setting out the detailed procedures and conditions for issuing short-stay visas for visits to the Schengen Area and airport transit visas. In February 2020, the New Visa Code entered into force. Inter alia, the new code defines visa requirements as a potential bargaining chip in gaining collaboration from origin and transit countries for the readmission of third-country nationals illegally present on the territory of the EU (i.e., the “paradigm of conditionality”, cf. Nicolosi, 2020, p. 471). In short, if the origin and transit countries do not collaborate in in re-admitting their citizens illegally present on the territory of the EU, third countries could be “punished” with restrictive measures, in primis concerning the issuing of visas (Nicolosi, 2020).

3.1 Patterns of Short-Term Visas Issued by European States

The trends in granting short-term visas (STVs) by the various Schengen states does not provide any evidence concerning the existence of a North-South divide in control practices. If such a divide existed, we should expect the older Schengen members to have a strict line on granting STVs, providing fewer STVs and turning down more applications, and the Southern countries would be expected to adopt a more relaxed line, refusing fewer applications and granting more STVs. None of these assumptions has been confirmed (Finotelli & Sciortino, 2013).

First, the majority of STVs granted by EU countries are released by a mix of older and newer members, notably France (39,124,476), Germany (30,776,452), Italy (23,736,365), Spain (19,963,026) and Poland (13,419,190). Visa practices reflect a variety of geopolitical, economic (tourism and trade), and historical considerations, rather than different attitudes towards migration controls (see Table 2.1).

Table 2.1 Short-term visas granted by Schengen Member States by round of Schengen membership
Full size table

Furthermore, the available data show the existence of significant differences among Schengen states in terms of the percentage of applications for STVs that are turned down (see Table 2.2). Nevertheless, the differences are not between North and South, but rather East and West. High rates of rejection define the visa practices of Western European states, both Northern and Southern. As Finotelli and Sciortino have shown for France and Belgium, the high rate of rejection must be understood in the context of these countries’ special role in Africa, a continent for which rejection rates are systematically higher (Finotelli & Sciortino, 2013). More recently, both Malta and Portugal have significantly increased their visa rejection rates, likely in the context of the post-2015 migration “crisis”.

Table 2.2 Percentage of short-term visas not issued by round of Schengen membership, 2004–2020
Full size table

4 Return and Removal Policies: Failed Convergence and the Inconsistency of the North-South Divide Argument

The “return” (voluntary or forced) of unauthorised foreign residents to their country of residence (or transit) is an important tool for preventing the settlement of an unwanted flow of asylum seekers and migrants (Coutin, 2015; Lindberg & Khosravi, 2021). Sovereign powers have always used political power to modify the composition of populations, often resorting to deportation and expulsion (Lindberg & Khosravi, 2021). Across the developed world, a striking fact is that, despite ostensibly great institutional and infrastructural efforts to remove irregular migrants and “failed” asylum-seekers, the actual returns are rather limited (Lindberg & Khosravi, 2021).

The term “removal” is used at the EU level to refer to what is commonly defined (in national policies) as “deportation” (De Genova, 2002). The expression “forced return” is also understood as synonymous with “removal”, especially in the EU political and legal context. Regardless of the term used to define it, removal is a specific form of return policies and practices. At the EU level, the 2018 Return Directive (henceforth, RD or the Directive) defines return as “ […] the process of a third-country national going back – whether in voluntary compliance with an obligation to return or enforced – to his or her country of origin, or a country of transit in accordance with Community or bilateral readmission agreements or other arrangements, or another third country, to which the third-country national concerned voluntarily decides to return and in which he or she will be accepted”. Namely, the Directive distinguishes between voluntary departure and removal.Footnote 2 According to the same articles, EU Member States shall take all necessary measures to enforce the return decision when no period for voluntary departure has been granted or when the obligation to return has not been complied with within the period granted for voluntary departure.

Return policy is one of the most controversial components of the Common European Asylum System (De Bruycker et al., 2016; Carrera, 2016). The European Council has repeatedly stressed, with particular intensity after the 2015 refugee “crisis”, the need for increased supranational harmonisation (EuroMed Rights, 2021). However, despite several attempts to increase policy harmonisation and/or intergovernmental operational cooperation among Member States over time, these have remained low. Removal is still largely a preserve of each state (Giuffré, 2015; Carrera, 2016; De Bruycker et al., 2016; Lindberg & Khosravi, 2021).

The very first attempt to move toward greater harmonisation of state return and removal policies and practices at a supranational level occurred in 1995, with the Treaty of Amsterdam (TOA). The Treaty conferred express power to the European Community (EC, today the EU) to address the issue of “illegal immigration and illegal residence, including repatriation of illegal residents” (Article 63(3) (b) of the EC Treaty). Since the very beginning, any attempt to use such power has turned out to be an “enduring punctum dolens” (Giuffré, 2015, p. 284). In May 1999, the Justice and Home Affairs Council attempted to manage the controversy between the Commission and the Member States, by accepting the provision that the EU and its Member States share the responsibility of closing readmission agreements with third countries (Giuffré, 2015; Carrera, 2016).

Further efforts towards this policy harmonisation were undertaken at the Justice and Home Affairs Council on 28–29 November 2002 with the adoption of the Return Action Programme. The Programme seeks to enhance operational cooperation among the readmission practices of the Member States, for example, introducing the systematic exchange of information among Member States and common training programs for return officials. The goal of greater policy harmonisation has been rather left in the background (Cassarino, 2010; Giuffré, 2015; EuroMed Rights, 2021). The most substantial effort to implement the Return Action Program is the 2008 Return Directive. It clearly distinguishes voluntary from forced return, and it succeeds in developing common minimum standards and guidelines on return. These include: the principle of voluntary departure (i.e., a general rule that a “period for voluntary departure” should normally be granted to irregular TCNs); a minimum set of basic rights for irregularly staying migrants pending their removal, including access to basic health care and education for children, etc., and a limit on the use of coercive measures in connection with the removal, based on the principle of proportionality. Nevertheless, most of the other purposes of the Return Action Programme, which should have led to greater operational coordination among Member State practices, have remained unaddressed (Coleman, 2009; Cassarino, 2010; Giuffré, 2015).

In response to the growth of refugee inflows to Europe following the 2015 refugee “crisis”, the European Commission has put forward several new measures in the field of return, such as a Recast of the Return Directive (EuroMed Rights, 2021). This brings a series of changes to the 2008 Directive, amongst which the most important is a new connection between return and asylum policies. Namely, under proposed Article 8(6), states shall issue a return decision (voluntary departure or removal) immediately after adopting a decision ending a legal stay, including a decision refusing refugee or subsidiary protection status. Additional novelties are the introduction of some limits to the applicability of voluntary departure (draft Article 9(1)) and of a short time period (5 days maximum) for refused asylum seekers to lodge an appeal against a return decision (the RD does not regulate the time-limit for appealing return decisions).Footnote 3

The topic of return is also addressed by the New Pact on Migration and Asylum (“the Pact”), released in September 2020, together with five accompanying legislative proposals (Jakulevičienė, 2020; Moraru, 2020; Vedsted-Hansen, 2020; EuroMed Rights, 2021). One of its goals is increasing the returns of irregularly staying TCNs from the EU, by means of four main instruments. The first is the appointment of a Return Coordinator within the Commission DG HOME, supported by a Deputy Executive Director for Return within Frontex and a network of high-level representatives (Moraru, 2020). These new positions should contribute to enhancing coordination, cooperation and consistency among domestic return practices as well as providing for clear monitoring tasks – e.g., accessible appeals mechanisms, special protection for vulnerable groups and independent monitoring mechanisms during the return procedures (ibid).

The second instrument is the connection between asylum and return procedures, following on from the 2018 Recast Return Directive. According to articles 53 and 54 of the Asylum Procedure Regulation,Footnote 4 asylum application rejection should be issued within the same administrative act with a return decision, in both border and ordinary return procedures. In addition to that, articles 40(i) and 41a(5) of the Asylum Procedure Regulation link the detention of asylum seekers to pre-removal detention during border procedures (Moraru, 2020).

The third set of instruments envisaged is novel screening procedure and a mandatory return border procedure, to prevent unauthorised entry into the EU and accelerate returns (Jakulevičienė, 2020; Vedsted-Hansen, 2020). These are applied to both asylum seekers – requesting international protection at border crossing points without fulfilling entry conditions – and irregularly entering third-country nationals – i.e., apprehended in connection with unauthorised crossing of external borders, disembarked following Search and Rescue (SAR) operations (ibid).

Finally, the Pact introduces a novel instrument, the “return sponsorship”, as a form of solidarity cooperation among the Member States (Moraru, 2020; Milazzo, 2021). Under this new scheme, a Member State commits to support returns from another one (Art. 45(1) (b) of the Regulation on Asylum and Migration Management).Footnote 5 The scheme also implies logistical, financial, and counselling help (Art. 55) provided by the supporting Member State (ibid).

Both the 2018 Recast Directive and the 2020 Pact have received much criticism by, as they raise several concerns regarding: human rights violations, violation of the right to asylum and principle of non-refoulement and the measurement of “effectiveness” of returns (Moraru, 2020; EuroMed Rights, 2021). At the time of writing, the negotiations concerning the approval of the Recast Directive are still ongoing. The Council reached a partial agreement, but the European Parliament is working within its Committee on Civil Liberties, Justice and Home Affairs Committee on negotiating its position (ibid).

4.1 EU Return Policies: A North/South Divide?

Given the highly fragmented approaches to returns in the EU (De Bruycker et al., 2016), it is difficult to see a process of diffusion or policy convergence in the area of return policy. It is likewise difficult to ascertain the possible existence of a North-South divide. A main problem is the lack of reliable and sufficiently detailed comparative data. Although Eurostat has recently been publishing some data on returns, they are very uneven, making estimates difficult. Nevertheless, through analysing these data, we can draw some useful conclusions.

Relying on the scant available data, we do not observe any systematic evidence of a North-South divide. Starting from the number of TCNs ordered to leave, we can see in Table 2.3 that the situation has been relatively stable between 2011 and 2019. More specifically, until 2015, it followed an alternating pattern of increases and decreases, yet without any remarkable changes. In 2015, with the refugee “crisis”, many European states announced the intention to strengthen cooperation with third countries on the identification and readmission of returnees in order to increase their increase return rates (EuroMed Rights, 2021). Notwithstanding, the general trend is still far from impressive. The total orders to leave increased only slightly between 2014 and 2016 (from 472,555 to 486,150).

Table 2.3 TCNs expelled and orders to leave issued, 2011–2020
Full size table

A similar scenario emerges when comparing 2011 with 2019 – the two temporal extremes of the period under observation. The number of TCNs ordered to leave is higher in 2019 but only to a limited extent (5%). Unsurprisingly, in 2020, the number of removals decreased sharply. In short, the available data do not support the North-South divide hypothesis. The countries who issued the highest numbers of orders to leave in such a period were from both Northern and Southern Europe: France (916,310), Germany (455,225), Spain (445,275), Belgium (333,275) and Italy (280,760).

Similar considerations apply to the volumes of removals carried out by each EU state.Footnote 6 The overall number of removals for the EU 28 area does not reveal any impressive increase in the aftermath of the refugee “crisis”, growing only by 9% between 2014 and 2019 (from 69,712 to 76,259). In 2020, because of the COVID-19 pandemic, removal operations plummeted remarkably, decreasing by 45% compared to the previous year. However, the number of removals carried out in the EU 28 area had diminished already in 2019, by 10% (see Table 2.3). Again, the data on actual removals do not support a North-South cleavage. Among the top five countries in the number of enforced returns implemented between 2014 and 2020 (looking at the totals in Table 2.3), we see a mix of Northern and Southern countries: Germany (115,737), France (75,010), the United Kingdom (69,862), Spain (65,215), and Italy (31,230).

An overall measure of effectiveness would certainly be the ratio between the total number of TCNs expelled and the orders to leave issued. The availability of data allows only for measuring it very roughly: many orders to leave are issued years before the actual return. We have considered the time interval between 2014 and 2020, since we have data for both the total number of TCNs expelled and the orders to leave issued (see Table 2.4). Our estimates are consequently indicative. Some of the results, however, are worth discussing.

Table 2.4 Ratio between the total number of TCNs forcibly returned and the orders to leave issued, for the top 5 countries in the number of enforced returns, in the period 2014–2020
Full size table

First, none of the countries considered is particularly effective in carrying out the enforced return operations. Even Germany, which stands out for the absolute number of persons forcibly returned, actually manages to remove less than a quarter of the TCNs ordered to leave. Second, we do not find any evidence of a North-South dichotomy. The most “effective” states are Germany and the United Kingdom (both remove 24% of the TCNs ordered to leave), Spain (23%), Italy (16%) and France (11%). The existing differences among them, moreover, seem to reflect the composition of undocumented populations – some nationalities being much more difficult to return than others – rather than the effectiveness of control procedures and infrastructures. Across Europe, the same factors – the implementation challenges of many readmission agreements, the low levels of collaboration among sending and transit countries, states’ administrative and budgetary constraints, and the enduring strength of embedded liberalism – operate in favour of low structural effectiveness.

A final consideration is needed. We are aware that, within this broader trend of low effectiveness, substantial differences exist, at both the national level and the EU 28 area, in the percentages of third-country nationals effectively returned depending on nationalities. A relevant report issued by the European Parliament in 2020 and the results of a research carried out by the European Migration Network in the same year clearly show that, between 2009 and 2019, the top 5 nationalities considered as “easier” to be returned are (alternately yet constantly) Morocco, Ukraine, Albania, Afghanistan and Algeria (Crego & Clarós, 2020; Vogel, 2020). This trend suggests a more collaborative approach to return procedures on the part of these five countries, as well as the reality of the situations in some other countries preventing the enforcement of return decisions issued against their citizens (e.g., Syria). In spite of that, as we have shown along this section, even when considering the most successful cases like the ones mentioned above, the overall return rates and ratio TCNs expelled/orders to leave issued remain overall remarkably low.

5 Conclusions

In this chapter, we have provided an empirical critique of the implicit conceptual frame of debates on convergence and divergence among EU states regarding migration policies. The key assumption underlying such debates is that EU Member States are essentially independent units, with their own admission and control challenges. This widespread belief undoubtedly plays an important role in European migration policy making. It is, in fact, a main “category of practice” (Bourdieu, 1997). The problem is that, very often, such a category of practice is turned into a category of analysis, becoming a stumbling block to the interpretation of reality.

We suggest, on the contrary, that migration control policies in Western European states have always been highly interdependent. From the Huguenot crisis of the 1680s to the so-called Tamil “crisis” of the early 1980s to the Belarus-EU border crisis of 2021–22, it is simply impossible to understand the control dynamics based on the idea of a set of essentially independent units, subject to the super-imposed authority of the EU. We should consider European migration controls as an interdependent, yet politically highly segmented, system (Bommes, 2012). Hence, it is difficult to meaningfully compare the policies and their outcomes of each European state individually. Because they are part of a system, the differences and similarities among them can be understood only when looking at the role they play within it.

This systematic configuration emerges clearly in the case of visa policies. Initially using inter-governmental powers, and, subsequently, the supranational mechanisms at their disposal, a group of core, mainly Northern European states, has managed to progressively impose their migration control goals upon reluctant Southern European states. As a result, the original Northern model has become the widely accepted normative model across all European states today, formalised in the New Common Visa Code. Turning to policies to actual practices, we have seen in Sect. 2.3 that no significant North-South differences exist in terms of STVs granted. The same is true as far as the numbers of not issued STVs are concerned.

Existing differences between North and South Europe have to be read in light of this generally homogenous background. As a fact, the duty of monitoring Europe’s “external” borders is thrust upon only a few states, i.e., the Southern Mediterranean – and, increasingly, Eastern – ones. As the requirement to obtain a visa prior to arrival in Europe makes reaching Northern European countries by flight far more complex and costly for many TCNs, they often opt to enter Europe irregularly using Southern (or Eastern) states as entry gateways. Such a role implies high costs for Southern and Eastern states, both financially and organisationally (Giordano, 2015; Italian Ministry of Defence, 2022).

At the same time, Southern countries often tend to “turn a blind eye” (Giordano, 2015, p.25) toward irregular migrants who refuse to register and apply for asylum in the first country of arrival (Following the Dublin Regulation), as they prefer to ask for asylum in Northern Europe. Looking at the official data about asylum requests, we can see that between 2014 and 2016 Sweden hosted the highest number of refugees per capita (2359 per one hundred thousand inhabitants against 254 per one hundred thousand inhabitants in Italy, in 2016). Moreover, in absolute terms, Germany received one in three asylum requests of the total in Europe, in the same years (European Parliament, 2022; UNHCR, 2022).Footnote 7

As regards return policy, we have shown how the scenario is highly fragmented and atomised. Strong coordination could actually be very useful, for example, in exerting unified pressure on sending and transit countries. The actual interdependence is, however, quite low and the barriers to the development of joint efforts quite substantial. While the amount of available data is overall small, these seem to point to two important considerations. The first is that there is no evidence of a North-South cleavage (as seen in Sect. 2.5). The second is that a process that we could define as “converge” is at play. Namely, all EU states (also including the United Kingdom) have shown to be largely ineffective in removing unauthorised TCNs from national and European territory.