Cross-Border Police Cooperation and ‘Secondary Movements’. On Reconfigurations in Enforcing Differential Mobility Rights within the Spatial-Legal Schengen Space

In the context of the so-called ‘migration crisis’, besides the politically more contentious introduction of border controls, on intra-European borders member states responded to onward mobilities – so-called ‘secondary movements’ – through border-area controls, bilateral (fast-track) readmissions and increasingly through joint patrolling of main cross-border routes. This article sets out to reflect upon the ‘Schengen crisis’ not by discussing the introduction of border controls, but by focusing on ordinary means of enforcement through border-area policing (Article 23 Schengen Borders Code) and through instruments of police cooperation, such as through joint patrolling or bilateral readmissions. By scrutinising the legal regimes of these instruments, plural in both scale and temporality, this article contributes to reflecting upon the productive reconfigurations in times of ‘crisis’ of the EU order and its enforcement.


Convention (CISA)
are also the anniversaries of so-called ' compensatory measures', which were presented as a means for compensating any 'security deficit' deemed to result from the lifting of border controls within the Schengen space. 15 Amongst these measures are institutionalised cross-border police cooperation (hereinafter CBPC) -such as hot pursuit and cross-border surveillance -and the setting up of the Schengen Information System (SIS), 16 as well as common rules regulating immigration and asylum. Amongst the lastmentioned 17 we find the roots of the subsequent Dublin Convention and Dublin Regulation. 18 As the main principle, asylum applicants are required to make a single application for international protection and criteria were established to determine which member state, out of all contracting state parties, is responsible for its assessment. The criterion of 'first entry' has been the most invoked in determining responsibility (obligation), even though it is situated at the bottom of a hierarchy of criteria. It is backed up through hard evidence of a 'hit' in the Eurodac database. 19 While the geographical scope and legal foundation of the Dublin regime has changed over time, its essence remained despite demonstrated ineffectiveness and inefficiency. 20 The need for reform was also expressed in the 2016 Dublin IV recast proposal, 21 in which the Commission nevertheless reaffirmed the Dublin Regulation as a ' cornerstone' of the legislative acts regulating the Common European Asylum System (CEAS). CEAS was initiated in 1999 with the aim of harmonising reception conditions and setting procedural standards as well as common criteria to grant international protection status across all member states. Its legislative components have been reviewed in several reforms.
The Dublin Regulation therewith does not grant freedom of movement to persons seeking international protection in the EU. Onward mobilities from the country of 'first entry' to other member states have been problematised as unauthorised 'secondary movements'. Additionally, in the absence of a mutual recognition of protection statuses, freedom of movement and establishment is likewise not a right automatically attributed to persons to whom a protection status is granted. Whilst it has been pointed out that asylum seekers have for years not necessarily made their application in the first country of arrival and have frequently successfully 'self-relocated', 22 the response to these intra-European mobilities has notably shifted from being to some degree tolerated 23 to becoming a political priority: a type of mobility aimed to be countered through 14 23 Campesi summarises the matter succinctly: 'The fact that the [EU border] regime was largely built on a (more or less explicit) precarious balance between central and northern EU countries on the one side, and frontline EU countries on the other side a multiplicity of means and increasingly through punitive measures. 24 This is one example of how the Schengen space of 'freedom of movement' constitutes a spatial-legal EU order of differential and unequal mobility rights and obligations. Asylum policy is deemed to mirror weaknesses of the European integration process. 25 The Dublin IV recast proposal was never adopted under the co-legislative procedure, due to polarised positions in the Council of the European Union especially between countries mostly situated at the EU's outward borders, which applied pressure to overcome the 'first entry rule' (calling for a greater 'burden sharing'), and other countries opposing an interpretation of 'solidarity' and 'fair responsibility sharing' (Article 80 TFEU) which underpinned proposals such as that of an emergency relocation and re-distribution scheme for asylum applicants. Such unequal mobility rights and obligations also raise questions on the enforcement of this regulation together with the enforcement of core rights for citizens and migrants. It does so even more in the context of a political 'stalemate' in the reform of the Dublin Regulation. In this context, will discuss in this article the significance of the 2017 Commission's recommendation to enhance proportionate police checks in border areas and to strengthen police cooperation, proposed as a means to also counter 'secondary movements'. 26 This article 27 proceeds as follows. I first outline the placement of bilateral police cooperation and national prerogatives on internal security within the legal framework of the 'Area of Freedom Security and Justice' (AFSJ). I then provide a brief introduction to the problematisation in scholarship and in policy of a legal 'patchwork' for cross-border police cooperation (Section 2). The main part (Section 3) describes and examines ' crisis' responses: from unilateral to bilateral enforcement instruments, and to their EU endorsement. Attention will be focused upon the regulation for administrative fast-track readmissions (3.1) and upon those of border-area patrols with a focus on joint border-area policing (3.2). Subsequently I engage with criticalities in the enforcement of the rights of EU citizens and of third country nationals in the context of enforcing unequal mobility rights through these instruments (3.3). I conclude by reflecting upon how these tools of enforcement could be allocated within different enforcement models of EU law and what this tells us about such enforcement (3.4). Section 4 elaborates in analytical terms upon the developments presented, by interrogating the role of temporality in the study of the law in the context of ' crisis'. Applied to a 'Schengen crisis' or its aftermath, how can we understand enforcements of 'the EU legal order' and challenges to integration, as well as scrutinise whether, and if so how, 'the past' dominates the present while at the same time the conditions for new things to become real are emerging? Following this logic, the article concludes (Section 5) with a reference to proposals on the EU legislative table.
As an analytical path, I propose to draw on anthropological scholarship in legal pluralism and to engage analytically with temporality. The article describes how in migration policing at internal borders and in cross-border police-cooperation multiple legal frameworks interact: they are plural in scale and policy fields, but also bear a particular rapport with temporality. The analysis enables the tracing of the ongoing reconfiguration of forms of power within what are understood as productive and generative effects of ' crisis': creatively combining pre-Schengen, Schengen and post-Schengen instruments. So-called ' old' and 'new' legal orders not only co-exist but also interact and are constitutive to 'the EU legal order' and its ongoing has also resulted in serious political weakness. The latter received financial compensations and technical support for acting as EU gatekeepers, while the former tolerated a certain degree of secondary movements of migrants not properly identified and fingerprinted according to the EU rules. reconfiguration. In my conclusion I suggest that it could well be the call for increased EU regulated 'policingmobilities' -such as the mobility of officers and of information -to emerge strengthened by the 'past crises' referred to in the opening of this article.

The AFSJ and state prerogatives
The Lisbon Treaty established, as an objective of the EU, the maintenance and development of the EU as an Area of Freedom, Security and Justice (AFSJ) -an area in which both the free movement of persons and the prevention and combating of crime are assured. Set out in Article 3(2) of the Treaty of the European Union (TEU), this objective now precedes the traditional goal of the internal market (Article 3(3)). 28 The AFSJ is regulated by Title V of the Treaty on the Functioning of the European Union (TFEU), 29 which covers (a) policies on border checks, asylum and immigration, (b) judicial cooperation in civil matters, (c) judicial cooperation in criminal matters and (d) police cooperation. Title V, however, does not affect member state responsibility for maintenance of law and order and internal security. 30 This primary competency also arises in the Schengen Borders Code (SBC). Whilst establishing that 'internal borders may be crossed at any point without a border check on persons, irrespective of their nationality, being carried out', 31 it also specifies that: (a) member states can temporarily introduce systematic border controls as a last resort in exceptional circumstances; (b) the absence of border controls does not affect member states' exercise of police powers to conduct spot checks and identity controls in border areas (regulated by Article 23 SBC). In addition, the Court of Justice of the European Union has no jurisdiction to review the validity or proportionality of operations carried out by the police or other law-enforcement services of a member state. 32

Cross-border police cooperation
Both the TFEU 33 and CISA 34 encourage bilateral police cooperation between member states, the latter in particular containing measures for countries sharing a common border through cross-border police cooperation (CBPC). Whilst informal cooperation across European border regions has been the rule rather than the exception, there has been an increased institutionalisation of CBPC, initially through bi-and multilateral agreements that often filled the initial legal void or grey area within which previous (experimental) practices had developed. In particular, the two main references for operational police cooperation in EU law today -CISA and the Prüm Decisions 35 -started out as intergovernmental agreements between a smaller number of contracting states and were only subsequently incorporated into EU law. Both set out that on sensitive matters (such as the attribution of executive powers beyond jurisdiction or temporal-spatial limitations in hot pursuit etc.) it is left to member states and their bilateral agreements to define the powers of cooperation in further detail. This constitutes a politically sensitive field, as formal police powers have traditionally been constrained within the borders of national territorial jurisdiction. Also within the legal order of the EU, new EU legislation on operational cooperation (and cooperation in criminal matters) follows the special legislative procedure (Article 89 TFEU). The regulatory framework for police cooperation in Europe has frequently been described by scholars as a 'fragmented legal terrain' or 'patchwork quilt'. 36  EU legislation in the field of the AFSJ has proliferated since Lisbon, without prior laws becoming obsolete, 37 arguing inter alia that ongoing law-generation processes not only provided recognition to regional cooperation initiatives but also produced legal uncertainty and conflicts between national and EU levels. 38 Most executive operational aspects are carried out by member states and specifically through horizontal bi-or multi-lateral CBPC. 39 At the policy level, particularly within the context of the recent ' crises' problematised by the Commissioner and introduced at the beginning of this article, CBPC regulation has become a topic of debate in several settings. In 2020 the European Commission argued that, contrary to other areas of the Schengen acquis, the legal framework for law enforcement cooperation has not been consolidated. It problematised scattered and inconsistent provisions in EU legislation (CISA, the Prüm Decisions, the Swedish Framework Decision etc.) as well as plural arrangements between member states constituting a complex web on a bi-or multilateral level. This situation is deemed to hamper effective law enforcement cooperation. 40 In its inception impact assessment concerning an EU Police Cooperation Code the Commission highlighted how some bi-or multilateral cooperation agreements provide for more extensive powers than EU legislation.

Plural legal orders and temporality
Bridging (socio-) legal police cooperation scholarship and legal anthropology, the analytical understanding of what is frequently described as a 'legal patchwork quilt', 41 can be enhanced by drawing on anthropological studies of legal pluralism. Legal pluralism has been proposed as an analytical framework or sensitising concept to approach the possibility that, within a geographic space or social order, different legal orders and bodies of law exist. 42 These orders are often of different scale and interact with each other (vertically, horizontally). Legal pluralism, rather than being a merely descriptive notion and classified as ' absence' of a legal framework, 43 captures a multiplicity of elements that de facto and de jure constitute police cooperation. Beyond the regulation of policing, it allows the investigation of interlegality 44 across policy fields, which matters when policing inner-European border areas: such as the interaction between laws regulating (a) national policing and police cooperation, (b) international protection and (c) those regulating the ' absence' of border controls at internal borders. An analysis of legal pluralism in anthropology or legal geography frequently puts, at the centre of empirical studies, 'law-as-practice' and dynamics of negotiations of plural legal orders by various situated or 'local agents'. 45 In this article, however, I would like to delve into details of law-as-text, highlighting specifically in Section 4 how legal pluralism and interlegality are useful analytical tools to engage with law and temporality in a 'post-crisis' situation from a social science perspective.

'Crisis' responses: from unilateral to bilateral enforcement instruments, to their EU endorsement
The 'hotspot approach' at the EU's external borders, set out in the European Agenda on Migration in 2015, 46 was the most publicised policy response to reinforce the Dublin system within Europe. Amongst multiple aims, it aimed to increase the fingerprinting rate of persons arriving irregularly in Greece and Italy, feeding into the Eurodac database. On intra-European borders, besides the politically more contentious unilateral introduction of border control, 47 member states responded to onward mobilities through border-area controls and especially initially through bilateral fast-track readmissions. In particular, from 2014 onwards, along certain routes and border sections, border-area patrols were intensified, readmission requests peaked and CBPC was strengthened. In September 2014 the heads of several national central border police directorates met in Germany to discuss possible bi-and multi-lateral responses to the 'migration emergency'. 48 Amongst the provisional agreements also figured 'the activation of bi-or trilateral joint control mechanisms along the main traffic arteries of interest to so-called secondary movements of migrants, who from South of Europe, especially from Italy, aim to reach countries of final destination'. 49 This was more than a simple ' activation'; as, for example, stated by a police union representative in a parliamentary hearing with regard to Italian-Austrian-German patrols, the joint controls (first set up in 2000/01) in this context had undergone a change in frequency and in intensity and were carried out within a modified trajectory, namely limited to the Italian territory. 50 Joint control measures, such as trilateral joint patrols on border-crossing trains set up by Germany, Austria and Italy, as well as by Germany, Austria and Hungary on a German initiative, 51 were also mentioned by the European Commission as best-practice examples in the Commission's Recommendation on proportionate police checks and police cooperation in the Schengen area. 52 This recommendation, issued in the context of ongoing border controls, is significant. It called for ' alternative measures' to border controls and recommended member states to shift towards and enhance proportionate police checks in border areas (regulated by Article 23 of the SBC) instead of border controls. Intensified policing of border areas was considered by the Commission to be a necessary and justified measure to counter what has been named together as three threats to 'public policy or internal security': (1) terrorism, (2) serious cross-border crime and (3) literally, 'risks of secondary movements of persons who have irregularly crossed the external borders'. 53 'Secondary movements', and even the mere risk thereof, were problematised as a security threat. In addition, there was a recommendation to enhance cross-border police cooperation through joint patrolling (hence the 'best practice' examples), as well as the application of bilateral readmission agreements. In particular, what had initially been national and then intergovernmental responses to unauthorised onward mobilities, became embraced on an EU-level: deemed suitable also for enforcing EU law, namely asylum policy and in particular the rationale of the Dublin Regulation, outlined in the introduction to this article.
Following this recommendation, scholars have examined the normative implications of enhanced borderarea measures with regard to Article 23 of the SBC. No scholarly attention has been given to what it means and 46 Further under n93ff. 47 Border control in 2015 was first carried out under art. 28 of the SBC (a situation requiring immediate action, emergency procedure, unilateral action through prior member state notification -maximum 2 months), followed by an application of art. 27 (foreseeable events, general procedure -maximum 6 months). Subsequently member states applied art. 29 (exceptional circumstances putting the overall functioning of the area without internal border control at risk as a result of persistent serious deficiencies relating to external border control as referred -maximum 2 years -dependent upon recommendation). implies when the likewise recommended police cooperation instruments (such as joint patrols) and bilateral readmission regimes are both enhanced and implemented for the scope of mobility control. Contrary to what some scholars suggested in the past, 54 bilateral readmission practices have not been substituted by the Dublin Regulation as an inner-European multi-lateral readmission agreement, and they have not become obsolete with the establishment of the internal market -neither in practice, nor in law.

Bilateral fast-track readmissions
In a publication on 'how police officers experience the refugee movement', an officer described fast-track readmissions across internal Schengen borders as follows: '[i]f in this job it is even possible to talk of an "experience of success", then it's those situations in which it is possible to intercept illegal immigrants quickly and to readmit them swiftly'. 55 According to statistics of the Police Directorate Tyrol -one of 9 Austrian federal Länder and workplace of this officer -the years 2014-2015 saw a peak in such swift returns, reaching 220-330 persons readmitted to Italy each year, before returns from Southern Germany into Tyrol following a denial of entry became more predominant.
Bilateral readmission agreements between member states that share a common border frequently contain a 'fast-track' or simplified procedure. Under specific conditions, a third country national who does not satisfy the criteria for entry or stay, can be returned within spatial-temporal limits to the neighbouring country, i.e. border region, if it is proved that the person entered from the neighbouring country and provided that several other criteria are fulfilled. A fast-track request that is denied can subsequently be resubmitted through the ordinary central channel. The simplified procedure falls under the jurisdiction of border-area police departments.
Whilst such readmissions are usually not listed amongst the instruments of cross-border police cooperation, historical as well as qualitative data of my research has shown how ' cooperation' features as a crucial and at times conflicting element between neighbouring police forces engaged in these practices. 56 Restricting my argument here to bilateral readmissions/agreements currently in force, I will outline how this instrument enabling a 'swift' return over an internal Schengen border (notably through a legally different regime from that of a denial of entry through border control) has its roots in a pre-Schengen political order and has been upheld in post-Schengen EU legislation, albeit lacking certain legal guarantees otherwise required for return procedures. Excluded from such 'swift' returns are persons registered in the Eurodac database, who fall instead under the scope of Dublin transfers. As the mobilities of asylum seekers increasingly shifted from being to some degree tolerated into becoming a political priority, fast-track readmissions gained renewed significance as a governmental measure to counteract so-called unauthorised 'secondary movements' of persons not registered in Eurodac.

Regulatory framework
The primary basis of this mobility control instrument is provided through bilateral agreements regulating 'the readmission of persons on [sic] the border', 57 as well as through related national law, setting for example the criteria and temporal conditions for administrative (pre-readmission) detention. Agreements between member states in force today mostly date back to the late 1990s, early 2000s, and a few to the 1960s. 58 Today, their ' anchoring' in EU law is found in Article 6(3) of the Return Directive. 59 Many of the agreements in force have notably been stipulated or renewed in the context of an envisaged abolition of border control throughout the Schengen area or during the enlargement processes. In 1997 the Schengen Executive Committee issued guiding principles 'for means of proof and indicative evidence within the framework of readmission agreements between Schengen States'. 60 Readmissions were mentioned in Article 23(4) of CISA but, according to the Committee, problems persisted between signatory parties in applying readmission agreements in practice. 61 The Committee aimed to ease in particular the temporal and evidentiary criteria contained as conditions for a successful readmission. 62 Agreements stipulated in the 1990s aimed to ameliorate the conditions for successful readmissions, by setting temporal and evidentiary criteria which were easier to satisfy. They also introduced stricter time limits for fast-track readmissions, compared with those found in agreements of the 1950s and 1960s. 63 This practice, in interaction with border-area checks, thus became more closely related to border-crossing interceptions in the border area. Provision is made for fast-track procedures today, for example, within 24 hours from an alleged border crossing in the Italian-Austrian agreement, 64 4 days in the German-Austrian agreement, 65 4 hours in the French-Spanish agreement, 66 and 24 hours or an interception within 10 kilometres in the Italian-Slovenian agreement. 67 Here we see (a) existing differences in the bilateral regulatory framework for these practices and (b) the strict time frame, indicating a relation to an irregular border crossing rather than to an irregular stay. A proposal to further regulate and 'harmonise' internal European readmission agreements, was part of the (never adopted) broader Finnish Initiative in 1999-2000. 68 The Return Directive 69 notably replaced the provisions of Articles 23 and 24 of CISA. It aimed to set ' common standards and procedures in Member States for returning illegally staying third-country nationals'. An explicit reference to bilateral readmission agreements between member states is contained in Article 6 (return decisions), namely as one of the ' exceptions' under which member states may refrain from issuing a return decision: 'if the third-country national concerned is taken back by another Member State under bilateral agreements or arrangements existing on the date of entry into force of this Directive' (Article 6(3)).
In the 2018 Commission's proposal for a Return Directive recast, 70 Article 6(3) was maintained. Within the co-legislative procedure, amendments were only proposed in terms of inserting the exact date: bilateral agreements in force prior to January 2009 are set to remain valid. Additionally, in the Commission's recommendation to enhance border-area policing, member states were advised that a 'proper' and ' efficient' application of the bilateral readmission agreements in accordance with Article 6(3) of the Return Directive  would be 'instrumental in addressing secondary movements of illegally staying third-country nationals'. 71 They are in practice also applied to counter irregular entries, not only stays, especially in the case of simplified fast-track readmission procedures following interceptions in the border area. As such, they interact with border-area police checks, regulated by Article 23 SBC which, in particular, cannot have border control as an objective. Mobile police patrols in border areas are a governmental technology which emerged within the neo-functionalist reasoning of ' compensatory measures' that accompanied the process of abolishing border control amongst Schengen members. 72

Joint patrols and border area control Article 23 SBC
In the context of increased Eurodac fingerprinting through the 'hotspot approach', fast-track readmissions have become less applicable to counter 'secondary movements', since persons have increasingly been registered in Eurodac and consequently, if intercepted in another country than that of registration, such person falls under the scope of Dublin transfers. It is in such a context that the deployment of joint patrols along exit routes should also be placed, tasked to impact (momentarily) on attempted onward mobilities before a person crosses the border into another EU member state. For the purpose of this contribution, I will concentrate on one specific form of joint patrols and their relation to Article 23 of the SBC, namely joint patrols set up along exit routes, set up -just as with the trilateral patrols mentioned above -with the explicit purpose of countering onward mobilities. According to a European Migration Network survey, 73 police checks carried out in border areas adjacent to internal Schengen borders are measures put in place by countries both for the purpose of crime control 74 and, in particular, for immigration-related checks. 75 This survey should be read in connection with the Commission's Recommendation on proportionate police checks and police cooperation in the Schengen area, 76 which recommended intensifying proportionate police checks in border areas and strengthening CBPC. In the course of 2014-2015 one can identify a tendency of member states to draw further on joint border-area patrolling experiences which were identified amongst the best practices outlined by the Commission in 2017. These included joint patrols along international railways connecting member states of so-called 'first arrival' and central EU member states.

Regulatory frameworks
The primary legal basis for joint patrols is found in the bilateral police cooperation agreements between the participating parties. Joint patrolling was not part of the Schengen instruments, but it emerged out of an additional initial multilateral convention outside the EU framework, namely the Prüm Convention (2005), 77 most of which was incorporated into EU law through the Prüm Decisions (2008). 78 While the Prüm Decisions are mostly known for regulating enhanced information exchange possibilities, Chapter 5 of Council Decision 2008/615/JHA covers ' other forms of cooperation', amongst which joint patrols feature as one type of joint operation. Legal aspects of relevance for operational police cooperation are the competencies and executive powers given to officers operating on another country's territory, as well as the use of equipment (such as weapons or vehicles) and the liability of officers. Depending on the cooperating partners, there are differences in the powers given bilaterally. The granting of executive powers to another country's police force either by (a) conferring executive powers of the hosting state and, in particular, by (b) allowing the carrying out of executive functions as attributed by the sending state represents a very sensitive topic, also expressed through the explicit reference to national legislation. 79 In addition, when carried out in border areas, joint patrols also interact with Article 23 SBC and the increased juridification of border-area policing, such as through interpretations of judgments of the Court of Justice of the European Union. Following up on the European Commission's 'Back to Schengen Roadmap' 80 and the 2017 Recommendation to enhance police controls in border areas and police cooperation, the recommended instruments became the topic of an ' orientation debate' within the Committee on Operational Cooperation (COSI) of the Council and were followed up by three meetings of the Law Enforcement Working Party (LEWP) of the Council. Member states were asked to outline their existing cooperation experiences and their legal base as well as being asked to respond to the question whether these instruments 'may constitute alternatives to temporary reintroductions of controls at internal borders'. 81 It is here that we find roots within the 'Schengen crisis' in which the enhancement of the current legal framework for police cooperation (mostly regulated through bilateral agreements in detail) emerged. In the concluding document, legal fragmentation was highlighted and it was argued that a 'more predictable regime' could be aimed for, 'where agreements on cross-border police cooperation are multilateral rather than bilateral, and possibly based on harmonized EU instruments rather than local agreements'. 82 Enhanced possibilities were welcomed, albeit not necessarily as alternatives to border control.

Reinforcement of differential freedom of movement and impacts on rights of EU citizens, persons with immigration status or seeking international protection
Whilst overall the search for ' alternatives' to border control is presented as a means to save the (economic) Schengen area and to guarantee the effective exercise of mobility rights, these measures, when applied to counter unauthorised mobilities, can impact on the enforcement of other rights of EU citizens and of third country nationals. (1) The Return Directive set procedural safeguards (decision in writing, effective remedy to appeal, etc.). The fast-track procedure, provided for in readmission agreements between member states that share a common border, does not provide for such safeguards within the terms of the bilateral agreements. In addition, when Article 6(3) is applied, the procedures continue to fall under the scope of the Directive. 83 Whilst fast-track readmissions have seldom been subject to academic analysis, in specific cases legal scholars and in particular legal advocacy groups have highlighted a breach of the right to legal assistance and judicial advice, and the related possibility of information, defence and appeal. 84 (2) According to the Return Directive, a bilateral readmission to another member state is followed up by the receiving country issuing an eventual return decision. However, it has also been argued that the Directive per se does not preclude the possibility of so-called ' chain readmissions', i.e. the receiving country in turn readmitting a person to another country from which the person had entered. 85 Chain readmissions, thus leading to the fast-track removal of a person across multiple countries, have more frequently been addressed by legal advocacy or in court, not least since chain readmission can potentially lead to a comparatively swift removal of a third-country national across the external EU border. (3) For readmissions under the Dublin regime, the jurisprudence of the European Court of Human Rights has set certain limitations by suspending temporarily 'Dublin transfers' of asylum seekers back to Greece 86 and of vulnerable asylum seekers to Italy. 87 Such considerations are outside the scope of fast-track readmissions (concerning persons not registered in Eurodac), even if these readmissions have been applied as a means to counter onward mobilities of persons aiming to make an asylum application in another country than that of first arrival. (4) Bilateral readmission agreements uphold the right to asylum and persons who make an asylum application can not be returned through such a readmission. This requires safeguards for the right to make an asylum application, even more so within the criticalities pointed out by legal advocates referred to under point (1). (5) Beyond these criticalities in regulation, both governmental sources and NGO reports have drawn attention to practices carried out outside the terms of bilateral readmission agreements, e.g. in the form of unilateral push-backs, group readmissions or a reluctance to consider individual cases and the situation of minors or vulnerable persons. (6) Readmissions are directly related to border-area controls, which are broader in scope and in effect. While the absence of border controls does not affect member states' exercise of police powers to conduct spot-checks and identity controls in border areas (regulated by Article 23 of the SBC), by definition, spot-checks in border areas cannot have 'border control as an objective' and are to be executed in a 'manner clearly distinct from systematic checks' (ibid). However, controls along exit routes aim to intervene not after the fact but on a future reality. By aiming to prevent or ' cancel out' a potential unauthorised bordercrossing, the border control objective is clear. Also, following the interpretation of the Court of Justice of the European Union, 'spot checks' need to be targeted and, when combined with targeting an 'immigration' offence, need to be based on police information and carried out selectively. 88 This puts citizens and migrants perceived to be immigrants -in what is a 'spot [and]check' -in a position of systematic targeting within the selection criteria, less able to be guaranteed not only freedom of mobility, but what is increasingly associated and confused in meaning: 'speedy' mobility. 89 7) Lastly and remarkably, it is the very European Commission, a key actor in the Schengen Evaluation and Monitoring Mechanism (responsible for monitoring compliance with the SBC) which is advising member states to make (intensified) use of their policing powers in border areas for the explicit goal of countering inter alia unauthorised 'secondary movements'.

Regulation and enforcement -asylum and police cooperation 3.4.1. Regimes for enforcing EU law
The editorial of a previous special edition of the Utrecht Law Review, engaging in discussions on mixed regulation and enforcement regimes in the EU, pinpointed three main levels of enforcement. 90 (1) Traditionally, enforcement of EU law has been the responsibility of national authorities and national supervisory authorities, with the clear influence of European law (e.g. national implementation and enforcement of EU law). Additionally, other trends have emerged: (2) the creation of networks of national authorities, aiming to coordinate and harmonise enforcement, by exerting (informal) influence on each other's; (3) an increasing shift not only from national enforcement to networks, but towards European enforcement through European supervisory authorities, with binding (or non-binding) powers. 91 Focusing on enforcement in different policy fields -from the financial to the environmental sector -so-called Europeanisation (see point (3)) has, for example, been analysed in the banking sector, where powers were transferred from the national to the European level through the creation of the European Central Bank. In the politically sensitive field of the AFSJ, EU agencies have gained an increasing role: whilst their operational activities are strongly linked with law enforcement on a national level, 92 it has been argued that asylum policy has also been increasingly enforced through such a 'verticalisation of enforcement'. 93 In scholarly studies of policing, EU agencies such as Europol or Frontex have been categorised as a form of 'vertical integration', opposed to 'horizontal integration' through bilateral cooperation.

Enforcement during the 'migration crisis'
When considered in the context of the so-called 'migration crisis' of 2015, the hotspot approach 94 was considered to be a measure that 'Europeanised' 95 the responses of Greece and Italy in dealing with arrivals at the EU's external borders. According to the Commission, it aimed to provide a 'platform for the agencies [seconded personnel of Europol, EASO, Frontex/EBGA, Eurojust] to intervene, rapidly and in an integrated manner, in frontline Member States when there is a crisis due to specific and disproportionate migratory pressure at their external borders'. 96 The deployment of seconded EU Agency personnel aimed to increase the fingerprinting rate and support subsequent measures, ranging from swift returns (Frontex) to enabling relocation (EASO). Registration and Dublin responsibilities remained those of the first EU country of arrival (except for relocated persons, the overall number of whom was small compared with the initial goals of the relocation programme). These measures, in which the objects of control were not just arriving migrants but also member states which had refrained from fingerprinting, 97 followed the 'indirect enforcement' through prior infringement procedures issued by the Commission against Italy and Greece (and Croatia). 98 At internal borders, besides the unilateral introduction of border control, we have seen a combination of national measures (border-area policing) and potential bilateral readmission of persons not registered in Eurodac, and on some borders a shift towards joint rather than mono-national patrolling of routes through bilateral police cooperation, also along exit routes and thus able to target and to impact additionally on the movement of persons already registered in Eurodac. Whilst EU Agencies have acquired competencies in the field of monitoring 'secondary movements', operational enforcement at internal borders has so far been carried out through intergovernmental coordination, in particular through bilateral police cooperation. This modality of enforcement may well be considered under the second modality of the 'networked' enforcement type described by de Cock Buning and others 99 as a means of exerting (informal) influence between cooperating partners (see point 3.4.1.), e.g. in readmissions or including through auxiliary (and not executive) functions in joint activities. However, talking broadly about inter-governmental or networked cooperation misses the detail, namely that, in practice, enforcement takes place in a decentralised way through the daily interaction between police officers working at the regional and local level. Considered within the political-historical context in which governments have been unable to find an agreement on the reform of the Dublin System, negotiating the enforcement of this controversial EU law is placed on the daily practices of those whom I have called 'street-level diplomats in uniform'. 100 What is thereby enforced is not only national immigration control, but also the EU Dublin regime. Whilst multiple measures have been adopted on the EU level, such as those outlined inter alia in the EU Agenda on Migration or in the New Pact on Migration and Asylum, one can identify not only an entanglement of enforcement regimes but also the predominance of bilateral cooperation at internal borders and its subsequent endorsement on the EU level. Whilst not leading yet to increased vertical integration, horizontal cooperation has been reinforced. As we will see in the concluding Section 5, the heterogeneity of CBPC in regulation and practice on a bilateral or multilateral level has been subsequently identified as a field in need of further EU regulation. Prior to outlining ongoing and upcoming reform proposals, it will be necessary to turn analytical attention to temporality, law and crisis.