The Rise and Fall of International Law in the Post-Lisbon AFSJ Legislation Cycles

The article focuses on the output and incidence of international law in the adopted AFSJ law-making for the period between 2009−14 and 2014−19, with particular emphasis upon asylum and immigration law. The article thus overall shows an initially rising but subsequently falling ‘international’ influence upon EU AFSJ directives and regulations. International law usage is significant even in times of populism or times of crisis-related law-making, particularly as to asylum and immigration law. However, the waning presence of international law also arguably indicates the development of the AFSJ as a booming legal field, where there is an operationalisation of a vast field of new actors, institutions, and systems through EU law. This account demonstrates how the EU shows a tangible intent to permit the influence of international law upon the AFSJ which supports well its general efforts to participate and engage as a global legal actor.


I. INTRODUCTION
The Area of Freedom, Security and Justice (AFSJ), which Article 3(2) TEU sets out as an 'area', has been gradually 'regularised' over time as a legal and institutional space and has a booming legislative agenda since the entry into force of the Treaty of Lisbon, estimated to be approximately 30% of the European Union's (EU) legislative output. 1 It has seen extraordinary advances and developments in a short period of time. The directives in the AFSJ are new legal tools that represent its regularisation in the post-Lisbon period with respect to the field of criminal and police cooperation, replacing pre-Lisbon Framework Decisions with direct effect. Regulations generally lack implementation requirements and obviate Member State's discretion and thus constitute a significant shift in new AFSJ law-making post-Lisbon. The precise pattern by which AFSJ law-making in this new era takes effect is evolving significantly already. For example, some express surprise that regulations have recently been adopted more frequently than directives in the AFSJ, thus warranting further exploration. 2 The metrics for assessing legislative output tend to involve subjective assessments as to their qualities and outcomes. Yet other means of evaluation of output might also seem desirable, employing the use of data and the lens of international law upon the AFSJ, given its rise, salience and significance to EU law-making, which this account develops.
Studies of authoritarian regimes and international law indicate that 'good' democratic regimes liberally use international law. 3 There are many legal and non-legal reasons for the EU's engagement with international law and international organisations, ranging from the EU's autonomous legal standing to the need to engage with international law if it is to be an effective actor. 4 EU openness to international law, particularly human rights instruments, is evident when it seeks to incorporate international law developments that are not binding upon it or required of it as metrics of good practice, developments which are accordingly studied here Some distinguish between instruments binding on all or some EU Member States, and instruments that are non-binding in nature but have been agreed upon in multilateral fora where the EU or its Member States (or both) are represented. 5 What is less clear is the extent to which the legislature incorporates these norms out of a sense of obligation and respect for international law stemming from Articles 3(5) and 21 TEU in the ordinary internal law-making. It is uncontroversial to state that the EU increasingly often helps shape, either directly or indirectly, developments at the international level and therefore has an interest in accepting international norms it had a hand in developing. In this regard, it is shown in this account how references to international law in preambles to legislation are arguably revealing of the EU's international ambitions and objectives in an area and informative of the EU's broader ambitions. Where they are not particularly revealing of objectives, they are arguably indicative of the EU's general commitment to international law. Studying the use of international law in legislative cycles is thus revealing of broader trends. Where no international law is used is also shown to be of interest in understanding law-making cycles and the EU's commitments to and engagements with international law.
Most references to international law are used in the preamble recitals of directives and regulations, emphasising their placement as legitimating, standard-setting, or justificatory. Although they cannot take precedence over those substantive provisions, recitals can help to establish the purpose of a provision 6 or its scope. 7 The Court of Justice of the European Union (CJEU) has stated that: 'Whilst a recital in the preamble to a regulation may cast light on the interpretation to be given to a legal rule, it cannot in itself constitute such a rule.' 8 The Court has also noted that: 'the preamble to a [Union] act has no binding legal force and cannot be relied on as a ground for derogating from the actual provisions of the act in question or for interpreting those provisions in a manner clearly contrary to their wording […]'. 9 Similar to many legal orders, recitals are used as interpretative tools in the EU legal order (as in legal orders in general for that matter) that the CJEU refers to in a restrictive manner. Recitals can help to explain the purpose and intent behind a normative instrument. They can additionally be taken into account to resolve ambiguities in the legislative provisions to which they relate. 10 This article analyses the use of international law in recitals in all adopted AFSJ legislation and focuses in particular on the area of asylum and immigration law over two full legislative cycles in current times to better understand output. The article thus compares directives and regulations in this period as the two main legal instruments used in these cycles. The focus here is on law-making from a legal perspective, examining the outcome of AFSJ law-making and international agreement sources used in the recitals of the preambles of adopted instruments focussing upon 1) broader AFSJ trends and then 2) trends in the sectoral field of asylum and immigration law.
This paper explores the incidence how the EU uses international law across a large-scale. 11 It does this through the preamble of regulations and directives because they act as a key focal point for use. The paper also considers the EU's participation or relations to the international law instrument used. The informalisation of EU external migration law is one of the most significant developments of recent times, where soft law measures are controversially used, from the EU-Turkey Statement to the Joint Way Forward on Afghanistan, 12 with the result of putting measures beyond judicial review and beyond institutional checks, with significant questions for the rule of law, compliance with international law and human rights. 13 As a result, the use of international law in internal EU asylum and immigration law becomes of much more salience, as much as AFSJ law, particularly where it is not per se required, in charting engagement with, for example, good law-making practice.
The article further gives context to the legislative cycle by examining the place of the EU in international organisations relating to the international law that it cites in its laws. It seeks to make this connection because the EU has increasingly appeared in recent years to constitute a distinctive global actor, as one of the last 'internationalists' standing. 14 The EU is committed in Article 21 TEU to being 'internationalist' as a matter of law, which is commonly (legally) understood to mean strictly adhering to international law, multilateralism, and promotion of 'good' global governance. However, the EU treaties give notoriously little guidance on how international law should be applied within the EU legal order. 15 As a result, legal scholarship usually considers the balance of how 'open' or 'friendly' the CJEU is to international law relative to the legislature. 16 However, it seems timely to evaluate metrics of international law within EU law-making across institutional actors focussing, for instance, on the legislature. The article thus considers the context of international law being used by the EU legislature internally by also examining the external role of the EU in international organisations and instruments externally. Sector-specific literature on EU legislative practice with regard to international law appears understudied as to the AFSJ general and specific output e.g. in asylum and immigration law over two cycles post-Lisbon. 17 It can thus contribute to the emerging literature on the role of the 'external in EU internal law-making. 18 This article will show how the use of international law rose or increased steadily in first-cycle regulations but fell suddenly in the second cycle. The account will show that these trends are largely borne out the area of asylum and immigration law. The article thus overall shows an initially rising but subsequently falling international law influence upon EU AFSJ directives and regulations. Although it attempts to expose a reduction in the international law influence in AFSJ law-making, it is not a uniform narrative. On the one hand, international law usage appears significant even in times of populism or times of crisis-related law-making, particularly as to asylum and immigration law. The assumption that in such domains there might accordingly be less open to international law is not per se accurate, as this account will show. However, the waning presence of international law also arguably indicates the development of the AFSJ as a booming legal field, where there is an operationalisation of a vast field of new actors, institutions, and systems through EU law. Here, new 'autonomous' EU systems and actors are key, not necessarily mimicking or downloading international systems, instead advancing a highly sophisticated yet far-reaching EU direction. 19 International law assists in tracing the nuances of this unfolding law-making.
Section II tracks 'On Numbers: Adopted Regulations and Directives 2009−14; 2014−19' and outlines the rising number of regulations, overtaking directives and their content. Section III examines 'On Areas of Law and Adopted Legislation and International law', setting out shifts per cycle and per instrument, Section IV discusses the 'Themes of Rising and Declining International law and internationalisation within Directives', outlining broad themes arising in directives and international law across the two cycles, and Section V discusses 'EU participation in international law of the AFSJ: on the "outside-in" and the "inside-out"' considering the place of the EU in international agreements and entities referenced in EU law, followed by Conclusions. The account thus provides overall trends and focuses upon both broader AFSJ and also more sectoral trends in asylum and immigration law with respect to the EU's role in the international law cited. The Eur-lex Legislation Directory was employed as the primary research tool for the time periods from 1 December 2009 to 1 July 2014 and 1 July 2014 to 1 July 2019. This time period was delimited to capture the first and second full post-Lisbon legislative cycle using legal basis, outlined below in detail. 20 Overall, nineteen AFSJ directives were adopted in the first post-Lisbon cycle, and 35 AFSJ regulations were adopted in the same period. Twenty AFSJ directives and 62 AFSJ regulations were adopted in the second post-Lisbon legislative cycle, from 2014 to 2019, in which regulations again overtook directives as the main instrument of the AFSJ legislative cycles. From the numbers above, the number of Directives has not actually decreased, but that there was an overall increase in legal instruments as a whole in the second cycle, the access of which has been occupied by Regulations. This regularisation of the domain through this specific instrument (i.e. a regulation) is notable in a legal field marred by competence issues unequally resting between the EU and Member States. A regulation arguably amounts to a clearer line to negotiate legally, unlike a directive, which can attempt for either minimum or maximum harmonisation and includes national discretion. During recent decades, EU Member States have increased their efforts in the fight against terrorism, organised crime, and illegal immigration, all of which have gradually developed a cross-border dimension, accentuated by the dismantling of internal border controls. Many recent AFSJ legislative developments appear to bypass the national security exception from EU competence (Article 4(2) TFEU) through legislating for inter-operability of new EU databases. The AFSJ has been swiftly integrated into transnational governance in times of crisis. 21 This shift thus represents a very clear direction to the EU's legislative outputs over the two cycles and is a significant development in studying law-making trends. One other explanation of output is that several potentially significant directives were not adopted because of political deadlock, particularly as to asylum and immigration law, but this is not explored further here for reasons of space and capacity to engage with its rationale. Overall, there were 67 (33 in directives, 40 in regulations) references to international law in adopted legislation in the 2009−14 and 2014−19 AFSJ legislative cycles. However, much context needs to be placed on such figures. There were 7 directives and 9 regulations in asylum and immigration law adopted in the first legislative cycle 2009-2014 and 2 directives and 26 regulations adopted in the second legislative cycle 2014-2019, thereby clearly mirroring overall trends outlined above as to the rising place of regulations in the AFSJ.

II. ON NUMBERS: ADOPTED AFSJ DIRECTIVES AND REGULATIONS
What can be said here firstly as to broader AFSJ cycle trends is that the tendency to invoke international law towards the end of the first legislative cycle also increased. Hence, there was a distinct rise in the number of regulations throughout the legislative cycle and a corresponding rise in the number of references to international law. However, as the above data demonstrates, there has been a dramatic drop-off in the amount of international law cited in the AFSJ Regulations in the second legislative cycle. One means to explain this trend is to consider that it is noteworthy that regulations overtook directives as the main instrument of use in the second cycle. It will be shown here that regulations are increasingly used to create autonomous EU concepts (e.g. actors, programmes, agencies, etc.) in order to operationalise the AFSJ without reference to international law. This accentuates the EU's own autonomous agenda, not showing any dependency or necessity to engage with international law, shifting significantly throughout the cycle more broadly but also more specifically in the area of asylum and immigration law. This development is analysed further in Section IV.

A. Areas of Adopted AFSJ Directives in the 2009−14 cycle
The AFSJ directives of the 2009−2014 legislative cycle can be loosely and informally grouped into distinct themes -namely, accused and victims' rights, fighting serious crime/terrorism, third-country nationals/asylum, and immigration and data-transfer related areas, relating to the broader official fields of human mobility and protection of borders, Common Asylum Policy, EU Migration Policy and Criminal Justice policy. Matters related to third-country nationals are the most common, 22 whilst the accused and victims' rights agenda forms a close second thereto. 23 Their concentration indicates the breadth of the legislative agenda in newer areas of legal competence as well as its political salience. The selection of directives over regulations in these fields indicates a concern with Member State's autonomy as to matters traditionally at the heart of national sovereignty, for example, procedural and substantive criminal law. It can also be stated that Articles 82(2) and 83 TFEU only allow for directives to be used as a legal basis. 24 Overall, as indicated above, the number of AFSJ directives increased throughout the cycle, similarly to regulations, but not to the same degree. A significant increase in the number of asylum and immigration law directives and regulations appears at the very end of the legislative cycle, also steadily increasing in number by year, consistent with temporal data on the EU's refugee 'crisis'although absolute precision about this time period is complex and highly contestable. 25 It also has a simpler example -a lot of legislation at the end of the cycle is explained by the fact that it takes several years for legislation to get adopted and political momentum to adopt it before the end thereof.
The  30 In the area of asylum and immigration law in the 2009-2014 legislative cycle, there were 7 Directives (using a broad diversity of legal basis of 78(2), 78(2)(d), 78(2)(f), 79(2), 82(2), 83(1) TFEU) as to trafficking, qualifications for asylum, third country national legal workers, international protection procedure, reception (asylum), third country nationals (corporate transfers and season workers), i.e. regular migration, thus using a broad spectrum of legal tools. In the 7 Directives, there were 12 references to international law: referencing many broadly accepted international instruments, i.e. with broad membership amongst the Member States and internationally, i.e. the Geneva Convention, UN Convention on the Rights of the Child and International Labour Organisation (ILO), Convention No. 29 Forced/ Compulsory Labour. There were thus few asylum and immigration directives, mainly as to irregular migration and few references to international law.

B. Areas of Adopted AFSJ Directives in the 2014−19 cycle
The directives of the 2014−19 legislative cycle can similarly be grouped loosely or informally into distinct themes, namely, accused and victims' rights; fighting serial crime/terrorism; third-country nationals/asylum and immigration. Matters relating to crime and terrorism are the most numerous, followed by accused and victims' rights, and the shift away from the dominance of third-country nationals/asylum/immigration is particularly noticeable. However, there is a noticeable upswing also in 'law and order' legislation as the EU battled perceptions of increasing crime, and justice-and terrorismrelated 'crises', and turned more towards preventive security. 31 The UN Convention on the Elimination of all Forms of Discrimination against Women is used in Directive 2012/29 to establish minimum standards on the rights, support, and protection of victims of crime. 32 Here, we witness an effort to broadly embed the salience of international law across EU AFSJ directives in the EU's first Criminal Law Directive (Directive 2011/36 on trafficking) and also the EU's pioneering victims' rights agenda. 33 To a similar effect, one of the directives citing the most international law is the Directive on the Presumption of Innocence (Directive 2016/343), citing the International Covenant on Civil and Political Rights, Universal Declaration of Human Rights, and UN Convention against Torture. It arguably provides evidence of the EU 'downloading' significant international law that its Member States are all party to in innovative law-making on the victims of crime, explored further in Section IV.
There is also a continuation of the operationalisation of the AFSJ in the second cycle with the development of many new EU actors, systems, and institutions. A small number of recent directives indicate a much more minor role for systematisation of a policy field through directives rather than regulations. For example, Directive 2019/884 notably contains no international law despite the development of a European Criminal Records Information System (ECRIS-TCN) for third-country nationals. To achieve a similar effect, Directive 2019/997 establishes an EU Emergency Travel Document System, without any reference to international law. The highly procedural nature of such directives, perhaps similar to other directives adopted later, signifies the evolution of the 'form' of directives in the AFSJ. As outlined above in the introduction, this arguably constitutes evidence of the EU developing autonomous actors and concepts, whilst granting Member States leeway in the development of the AFSJ. Ultimately, however, the two recent AFSJ cycles testify to a broader usage of regulations across the board and directives as more exceptional instruments.
In 2014-2019, there were very few directives, with a considerable drop from the already limited number in the first cycle, with 2 Directives in the area of asylum, immigration and border control (using the legal basis of: Articles 79 and 82(1) TFEU) as to third country national (entry/ residence for research, education), exchange information third country national criminal records (ECRIS). In the 2 Directives, there were zero references to international law as outlined above. This decline will be considered further in Section IV.

C. Areas of Adopted AFSJ Regulations in the 2009−14 cycles
Regulations in this cycle are dominated by what can be informally labelled systems, borders, visas, civil procedures, criminal justice, and third-country nationals. There was a significant emphasis on asylum and immigration law-related matters in the first-cycle regulations. In several asylum and immigration law regulations of 2010 relating to long-stay visas, a European Asylum Support Office (EASO) and third-country national visa exemptions at a key point in the EU migration crisis, there is only 1 reference to leading international law agreements, i.e. the Geneva Convention in the EASO Regulation. 34 This omission is surprising given that the instruments span a vast array of policy fields with no significant international law underpinnings therein: residence permits, visas, visa exemptions by country, or border checks.
To a similar effect, no references to international law are found in the 2011 Regulations, for example, on a European Network of Immigration Liaison Officers, an External Borders Agency (Regulation  1168/ 38 There was a continuation of regulations being used for AFSJ operationalisation, for example, Regulation 514/2014 on an Asylum, Migration and Integration Fund. Notably, the regulation using the second-largest number of international law instruments was the one establishing the European Border Surveillance System (Eurosur), Regulation 1052/2013, followed by Regulation 656/2014, establishing rules for the surveillance of the external sea borders. 39 Given the challenges of seeking unity in EU external migration law generally, a more regularised use of a 'uniform' EU legal instrument here such as a regulation is of note. 40 This cycle arguably shows efforts to seek external legitimisation of law-making in a contentious subject area developing a diverse portfolio of actors.

D. Areas of Adopted AFSJ Regulations in the 2014−19 cycle
In the 2014-2019 cycle, 'systems' continue to dominate along with civil procedures, civil justice, thirdcountry nationals, and data as they are rolled out. However, the 'areas' of law and policy became difficult to decipher in this second cycle. For instance, two regulations were adopted in 2019, creating interoperability across six AFSJ databases, and thus the broadening and widening of the AFSJ substantive and procedural areas evolve through such cross-cutting instruments. 42 Two 'cross-cutting' regulations were adopted on the interoperability of AFSJ information systems in 2019, Regulation 2019/817 EU-LISA and Regulation 2019/818, to check data on individuals whose names were stored on six EU databases (VIS, SIS II, Eurodac, EES, ETIAS, ECRIS ITN). 43 The number of adopted regulations being used to operationalise the AFSJ through the development of new actors, systems, agencies, and programmes relating to asylum, immigration law, for example, a European Coast and Border Guard, EU-LISA, and ETIAS-TCN, increased steadily in the second cycle of 2014−19. 44 This trend continues to occur through regulations − unlike directives, which give discretion to Member States. It arguably serves to emphasise the operationalisation of the broader legislative agenda taking place at the EU level. Regulations from this cycle mostly do not make explicit reference to international law despite many questions of fundamental rights raised by these systems with their surveillance agenda. It can be suggested that reference is implied more than explicit but this does not appear to be the case in the view of the author overall given the prevalence of autonomous concepts.
A small minority of regulations cited international law exceptionally and in abundance, in regulations relating to agencies: Regulation 2016/1624 (European Border and Coast Guard), and Regulation 2016/1625 (European Maritime Safety Agency). 45 There is arguably evidence of attempted 'legitimation' actions here on the part of the EU by using international law in these instances, for example, Regulation 2016/1625 referencing all nine key international law instruments relating to the law of the sea and asylum and immigration law. This is considered further below in Section IV. 46 In 2017 alone, eight out of ten regulations related to asylum and immigration law and thus showed continuity with the first post-Lisbon legislative cycle thematically, although all eight were without reference to international law. 47 One of these, Regulation 2017/458 amended the EU borders code to provide for additional checks on EU and non-EU citizens at borders, citing no international law provisions per se but referencing Interpol's lost and stolen databases. 48 The shift towards regulations continues to be distinctive where regulations enable no Member State's implementation or discretion in such an instrument, which is overwhelmingly used for EU 'institution building', 'actor establishment', or 'systems establishment'. More significantly, as developed in the introduction, no international law is used or cited in regulations to establish entities as part of a 'systematisation' or systems establishment of an asylum, immigration and borders control field (e.g. Coast Guard, etc.) as autonomous AFSJ concepts emerging in EU law. This constitutes evidence of the EU developing autonomous actors and concepts even in sensitive fields. This operationalisation includes powers to post immigration liaison officers to non-EU countries and to extend the scope of EU operations there (e.g. through Regulation 2019/1240). 49 However, proliferation of autonomous concepts can also be said to align with certain There were thus significantly more regulations than directives here, a notable trend in the two cycles of the AFSJ.

INTERNATIONALISATION WITHIN THE AREA OF AFSJ
Although the focus of this paper is predominantly on shifts as to Regulations in law-making, this is not to suggest that Directives have become inconsequential. Directives are important new legal tools in the AFSJ because of the absence of Member States' discretion in their implementation in an area with complex links to the States' competences and powers. Their demise in adopted legislation is thus significant in times of crisis and populism, as the first and second legislative cycles have experienced, particularly linked to debates on asylum and immigration law. A select number of themes are considered here as a means to synthesise the law-making period, which are inevitably arbitrary but are chosen nonetheless as representative means of reflection. The examples chosen below mostly reflect upon the AFSJ with application to asylum and immigration law also as far as possible.

A. Rising international law in AFSJ lawmaking
There are three broad themes considered here as evident in the initially rising international law trends in AFSJ 2009-2014 directives, where international law is used with greater incidence and evidence, namely: i) legitimation, ii) standard-setting, and iii) innovation in EU law-making.

(i) International law as legitimation of law-making
First, it might be said that permitting international law in directives results from practical necessity, where a law-maker seeks international law with both normative and practical significance, thus adding a certain legitimation to the law-making. The last year of the EU's 2009−14 cycle demonstrates the increasing salience of asylum and immigration law to the legislative agenda, in which almost half of the AFSJ directives relate to asylum and immigration law. 50 The instruments deployed demonstrate EU's efforts to enable global influence in its laws through international law, even in a difficult policy cycle. 51 Here, the EU has acted as an emerging global soft power grappling with a complex refugee crisis and has sought to 'download' international law for legitimation purposes. The high concentration of international law in key instruments is also notable, where the quantity of references used is 'bunched together' and thus high, for example, citing all major international law instruments relating to the law

(ii) EU usage of international law as standard-setting
Second, permitting global influence can be viewed as the engagement in standard-setting, showing the EU seeking to act as a strong internationalist institution. In an example of broader AFSJ law-making trends, for example, the UN Convention on the Protection of the Rights of the Child forms the main external norm or source for maximum harmonisation in Directive 2011/92 on combating sexual abuse and sexual exploitation of children and child pornography, to which the EU is not a party. The Convention is the most widely ratified of all human rights treaties and explicitly seeks full and direct incorporation to maximise its full legal effect. 52 As outlined above, the EU has ratification of the Convention as a strategic agenda and places emphasis upon an internal and external interlinkage of policy. It adopts a maximum harmonisation approach in Directive 2011/92, with far-reaching criminal law penalties, in Article 4(5) of it. This raises questions as to subsidiarity because it differs significantly from minimum harmonisation more usually found in AFSJ directives and thus attempts maximum norm convergence therein, not yet raised formally to the best of the knowledge of this author. This also raises the link between asylum, immigration and securitisation. 53 Here, the Convention has an important legitimising effect on EU law-making in this field. Explicit 'norm-taking' here legitimises EU's practices yet also demonstrates the EU's acting as a 'standard-setter' rather than just as a 'normfollower.'

(iii) Downloading of international law in innovative law-making
The explicit invocation of international law sometimes manifests in 'downloading' of norms, particularly in innovative areas of AFSJ law-making, meaning that it is used and circulated in the EU legal order. Apart from the Geneva Convention, the UN Convention on the Protection of the Rights of the Child forms the main external norm in the first EU Criminal Law Directive 2011/36 on trafficking, 54 as well as in Directive 2013/32 on common procedures for granting and withdrawing international protection, 55 Directive 2011/92 on combating sexual abuse and sexual exploitation of children and child pornography, 56 Directive 2013/33 laying down standards for the reception of applications for international protection 57 and Directive 2011/95 on standards for the qualification of third-country nationals or stateless persons as beneficiaries of international protection. 58 They give an insight into how elements of the EU agenda converge internally and externally.
Directive 2011/36 is highly significant in its objectives for preventing and combating trafficking in human beings and protecting its victims and a good example of the development of asylum and immigration law here and is also AFSJ best practice. It uses the UN Convention on the Rights of the Child, the UN Convention on Transnational Organized Crime and Protocols thereto, and the UN Convention relating to the Status of Refugees, and is an important strategic use of international law by the EU given longer-term EU policy in this area. 59 However, some have argued that the Directive was considerably watered down in the end so as to allow national preferences to be exercised. 60 It has also been argued that the Directive lent itself towards implementing security maintenance rather than victim protection and rehabilitation, a concern applicable more broadly to EU AFSJ law relating to asylum and immigration. 61 Still, it is considered a landmark point in EU law and is a vivid example of the downloading of international law in AFSJ law-making.

B. Declining international law in AFSJ law-making
The legislative output of the two cycles broadly supports the thesis of their being declining international law usage in AFSJ law-making. There are three broad factors evident in declining international law in AFSJ 2014−19 Directives: i) the rise of Regulations, ii) multiple actors developed in Regulations and iii) multiple systems developed in Regulations.

(i) The Rise of Regulations
The operationalisation of the regulation as an AFSJ instrument is an important continuing trend in an area of law-making sensitivity of the Member States. This is particularly so in the areas of asylum and immigration law. The number of regulations used to operationalise the AFSJ through the development of new or revitalised agencies, actors, systems, and programmes increased dramatically, particularly relating to third-country nationals, but also generally as to packages of law-making as to asylum and immigration. An absence of reference to international law is found in all regulations to establish new autonomous AFSJ concepts, in other words, entities forming part of a systematisation of a policy field (e.g. EPPO, Coast Guard). Certain 'new entities', such as the European Border and Coast Guard, as independent EU agencies represent a continuing trend towards agencification of the AFSJ. Their evolution is unlinked to international activities or programmes. They are thus autonomous developments in AFSJ law-making, affording less discretion to Member States to implement the AFSJ. It thus signifies its growth as a booming legal field.

(ii) Multiple New Actors in Regulations
Some 'new' actors developed as a result of the AFSJ, such as the European Border and Coast Guard, have in fact evolved from existing agencies (i.e. Frontex) and represent over 30 years of evolution of the Schengen Agreement (Regulation 2016/1624). 62 In 2017, a regulation was similarly used to establish multiple actors without reference to international law, such as the European Public Prosecutor's Office (EPPO) (Regulation 2017/1939) 63 and the aforementioned European Border and Coast Guard, comprising a vast array of measures from 'hot spots' to travel documents. The European Border and Coast Guard will eventually have 10,000 operational staff, representing a significant increase in EU capacities and transforming Frontex into a quasi-federal agency. 64 The EPPO was adopted after a very long and complex negotiation within the Council and a decision by 20 Member States to proceed through enhanced co-operation. As the centrepiece of a new system joining national law enforcement, judicial authorities, and EU actors such as the European Anti-Fraud Office (OLAF), Eurojust and Europol appear to constitute an example of a clearly autonomous EU concept arising to support the emergence of more complex investigations. Of course, all AFSJ agencies are generally thought to circumscribe the Commission's executive powers and to weaken the oversight powers of the European Parliament but conversely improve the influence of the Member States. 65 It is not a 'given' that a rise of AFSJ actors per se aligns with the rise of, for example, crisis or populism in this era.

(iii) Multiple Systems developed in Regulations
New systems adopted in the AFSJ recently (EES, 66 ETIAS, and ECRIS-TCN 67 ) represent the so-called 'third wave' of databases and a new generalisation of surveillance of movements of non-citizens based upon the concept of interoperability of systems. They move on from the 'first wave' systems of the Schengen Information System (SIS) and Eurodac, concerned with modernising immigration control and allocating responsibility amongst the states and the 'second-wave' databases associated with the 'War on Terror', the VIS database and SIS II, that were 'upgraded'. 68 Three regulations were adopted to extend the use of SIS II further in 2018. 69 ETIAS is a broad system for all third-country nationals entering the Schengen area, 70 which will be paired with a system for monitoring the presence on the Schengen territory of visa 'overstayers' ('EES' Entry/Exit system (Regulation 2017/2226)). The EES and ETIAS introduce surveillance for almost all travellers grounded through automaticity and blind reliance on technology. This appears overall to indicate continuity between the legislative cycles in terms of evolution. It also represents a significant push to institutionalise the AFSJ through far-reaching and controversial 'autonomous EU concepts' in sensitive fields traditionally close to Member State's sovereignty. They are not related to international law activities of the EU. It is notable that this third wave occurs through a regulation not a directive (i.e. without Member State discretion) and mostly without reference to international law. This suggests that the EU legislative cycle has found it easier to process regulations as the output of the cycle. This further indicates that the EU is not seeking sources of external legitimacy from international law in regulations because they are not deemed to be needed in an 'autonomous EU concept'. 71 For example, ECRIS-TCN is a new centralised system for the exchange of criminal records on convicted third-country nationals and stateless persons and is meant to complement the existing decentralised ECRIS system where information on criminal records of EU nationals is exchanged. 72 It appears to indicate continuity between the legislative cycles and a significant push to institutionalise the AFSJ through autonomous EU concepts. The European Data Protection Supervisor (EDPS) in Opinion 4/2018 warned that an interoperability regime created a new centralised dataset with information on millions of third-country nationals and the potential for a highly serious data breach. 73 It is also worth noting that no international law is referenced in the system despite an emerging default retention period for data of five years notwithstanding Opinion 1/15 EU−Canada PNR of the CJEU. 74 Perhaps it is the case that legislators see this as a continuation of past legislative practice without the need for external legitimation practices. This does not bode well given that no non-EU citizen will be left 'unsurveyed' through at least one database and the aggregation of data will not only generate new databases but also transform existing databases created originally for administrative, immigration control purposes into powerful tools. 75 This also indicates a very specific direction of the AFSJ through autonomous concepts without reference to international law that could have sought to use international norms as normative benchmarks or standards. The breadth of international law that could possibly be used is notable in all domains, particularly those relating to asylum and immigration law. This would have added legitimation or set out prescriptive concerns to follow best international rules, systems, or practices.

THE 'OUTSIDE-IN' AND THE 'INSIDE-OUT'
The next section briefly outlines the EU's participation in the international law of the AFSJ in the two cycles, focussing upon international law referenced in asylum, immigration and border control instruments only, as a means to understand the 'inside-out' 'outside-in' dynamic of international law emerging here. As noted above in the introduction, this casestudy takes a snapshot from this specific field of asylum and immigration law and border control as to overall trends sketched more broadly. As stated above also, there are many reasons for the EU's engagement with international law, ranging from the EU's autonomous legal standing to the need to engage with international law if it is to be an efficient actor. 76 Prior to the demise of the Third Pillar, Council of Europe instruments were often used as the primary source of international law in AFSJ law-making, along with UN instruments to a lesser extent. AFSJ legal instruments adopted in the 2009−14 legislative period appeared to show a preference for broadly 'accepted' international law. 77 This appears generally borne out in the two cycles accounted for here, although a fuller study of membership is beyond the scope of the present work.
The international law referenced in the two cycles of AFSJ legislation as to asylum and immigration law outlined above is as follows in full 78  The international law used generally consists of broadly accepted international law instruments of which all Member States are a party thereto mostly. 79 Overall, the international law sources used mostly demonstrate a commitment on the part of the EU to participate in international organisations and agreements actively and to engage with key international law relating to the AFSJ, even in times of crisis related law-making. The account thus briefly examines the relationship between the EU with the international agreement or organisation references in its AFSJ legislation.
For reasons of space, a selection is considered here as to the EU's role therein. This account selects four widely used in the legislative output of the two cycles of asylum and immigration law and considers the EU place therein, namely UNCLOS, UNHCR (qua agency), UN Convention on the Rights of the Child and the Palermo Protocol.

A. UNCLOS
The UN Convention on the Law of the Sea (UNCLOS) features regularly in EU legislation in the cycles outlined. The EU is a party to the UNCLOS but is not able to participate in all areas of the UNCLOS institutional system because for the EU to participate it must either amend the constitutive instrument 76 See above (n 4) and (n 16). 77 On AFSJ Directives, see Elaine Fahey, 'Joining the Dots: External norms, AFSJ Directives and the EU's Role in Global Legal Order' (2016) 41 ELRev 105, 109-11. 78 Multiple spellings and versions of abbreviations and instruments are used throughout the two legislative cycles in official legislation and the most common are listed here. 79 E.g. 'Geneva Convention' has multiple variations (with/ without years), without/ without UN, fuller description, etc.
of the organisation to join or participate without full membership. Despite exercising significant competences in the fields dealt with by the bodies, the EU is neither a member of the International Maritime Organisation (IMO), nor the International Civil Aviation Organisation (ICAO). The EU has joined a number of UN Conventions such as the UN Convention on the Rights of Persons with Disabilities, UN Convention against Corruption, the UN Convention against Transnational Organised Crime and the UN Framework Convention on Climate Change. The EU has long struggled with the reluctance of some Member States to accept the EU advancing its role in the transport sector in particular and it has long faced deadlock in some important UN bodies e.g. UNHCR. 80 At EU level, maritime activities involving security and police activities fall within the shared competences. The role of the EU in Search Rescue is problematic as the EU is not a member of the IMO and thus is not party to the relevant Conventions. Even if search and rescue activities (SAR) are meant to be included in the EU's maritime policies and coordinated at EU level, the existing institutional framework does not allow for such a development. However, the EU is party to the UNCLOS and yet a unified EU flag does not yet exist and this has watered down the EU's duty of assistance which falls on state flags. Yet, the EU has shared competences in the field of transport and has shown interest in navigation, creating different forums where stakeholders can exchange information and organise cooperation. 81 Thus, the inclusion of references to UNCLOS in AFSJ legislation is of note in line with its broader engagement policy.

B. UNHCR
The United Nations High Commissioner for Refugees (UNHCR) is also referenced through legislation in the two cycles and is a UN agency of note, cited more than any other body or organisation and is not thus an ordinary international law instrument. 82 The EU's engagement with the UNHCR dates back to the early 1990s when the latter established a liaison function in its Brussels office to monitor envelopments and input to the EU's emerging harmonisation process in the areas of asylum and immigration. In 2005, the UNHCR and the EU signed a Strategic Partnership Agreement to consolidate, develop and better structure the relationships on protection and assistance for refugees. UNHCR's office in Brussels started a process of networking UNHCR branch offices in national capitals through the establishment of EU focal points in these offices. These were charged with monitoring and influencing their government's positions in negotiations on EU asylum instruments.
As the UNHCR is an agency of the UN, the issue of membership does not arise as all EU Member States are members of the UN and the EU can influence its policies through its observer status. Ever since the Commission and Council started to develop the external dimension of EU asylum and migration policy, aimed at improved co-operation in the joint management of migratory flows, UNHCR has monitored this process closely and has provided expertise and policy inputs as regards EU cooperation with third countries (Eastern Europe, Western Balkans, Mediterranean basin) in asylum and migration matters. The support therefore in legislation is thus significant given the broader engagement ongoing.

C. UN Convention on the Protection of the Rights of the Child
The UN Convention on the Protection of the Rights of the Child from 1989 is used in much AFSJ legislation relating to asylum and immigration law in the two cycles outlined. The EU has ratified the Convention as a strategic agenda and places emphasis upon an internal and external interlinkage of policy, which is a Convention to which all new EU Member States must accede. 83 The Convention constitutes a landmark in international law by establishing children's status as legally empowered subjects of entitlement. Over the last two decades, the Convention has shaped how states and non-state actors think about children in legal, policy and normative terms at the societal level. It retains its appeal as the lingua franca of children's rights advocacy at the regional, national, European and international levels. 84 The EU is understood to have carved out a distinct and potentially extremely powerful role in developing and enforcing children's rights, not just across the Member States of the EU, but in non-EU states as well. The explicit EU constitutional reference to the protection of children's rights as one of the core objectives of the EU (Article 3(3) TEU) further reinforced by the increasingly explicit allusions to the Convention in the substance of binding EU legislation. 85 The Convention notably forms the main external norm or source for maximum harmonisation in Directive 2011/92 on combating sexual abuse and sexual exploitation of children and child pornography, to which the EU is not party, in its earliest AFSJ law-making. The Convention is the most widely ratified of all human rights treaties and explicitly seeks full and direct incorporation to maximise its full legal effect. 86 Here, the EU adopts a maximum harmonisation approach in Directive 2011/92, with far-reaching criminal law penalties, in Article 4(5) thus shows how it engages with the Convention in a direct and far-reaching way.

D. UN Protocol to Prevent, Suppress and Punish Trafficking in Persons (Palermo Protocol)
The prevention of Trafficking is an important law and policy domain showing strategic use of international law by the EU in line with the longer-term EU policy in this area. 87 The first international definition of trafficking in human beings was agreed in the 2000 UN Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially women and children (the Palermo Protocol). 88 Also, all European Union Member States with the exception of Ireland ratified the 2000 UN Protocol against Migrant Smuggling by Land, Sea and Air, which supplements the UN Convention against Transnational Organised Crime (UNTOC), as to the latter which the EU has joined. The EU became party to the Protocol in 2006, annexing a declaration to the Protocol with respect to the EU's and its Member States' competences the subject of the Protocol.
The Palermo Protocol is said to serve as a broad model in a variety of contexts. The EU has previously adopted the values of the UN Palermo Protocol within its treaties. 89 The Union's approach to the trafficking in human beings has explicitly sought to differentiate itself as 'holistic', 'multifaceted' and even multidisciplinary. 90 However, the EU was previously perceived as favouring the limitation of irregular migration and not ameliorating the situation of the trafficked. Nonetheless, at both international and European level, opposition has long-existed between a rights-based and law enforcement approach. The new Trafficking Directive was the culmination of more than a decade of developing EU strategy to combat illegal migration. 91 It is, however, notably perceived as an innovative instrument because of its character as a criminal law entity, thus taking a distinctive turn as to remedies, rights and enforcement. 92 The Directive also contains provisions which are largely similar to the Council of Europe Convention 83 on Action against Trafficking in Human Beings. 93 As Chaudary states, the Directive moves away from the previous EU law approach of subordinating protection measures to investigating and prosecuting human traffickers for acts already committed. 94 The Directive is thus understood to make a very explicit and specific use of the Trafficking Convention (e.g. in recital 9) in providing a 'solid' legal base upon which to successfully advance a trafficking case and is a clear example of adoption of international law in line with the EU's international policy. 95

VI. CONCLUSIONS
The article has shown how in the legislation adopted under the AFSJ over two full cycles, a broad variety of trends as to international law is evident and revealing because it provides insight into the EU's international law activities and its own autonomous development of AFSJ concepts. It develops debates on the place of international law in the EU legal order.
It has explored the openness of the AFSJ to international law, looking in detail at the area of asylum and immigration law. It thus focused upon the output and incidence of international law in adopted AFSJ law-making for the period between 2009−14 and 2014−19, with particular emphasis upon asylum and immigration law. It is the first and second fully regularised legislative cycle of the post-Lisbon era. Here, the AFSJ became normalised as an area of law institutionally, competence-wise, and enforcementwise. The article has outlined how law-making practices initially demonstrate substantial influence of international law on EU law. A decline in the reference to international law appears to indicate the further development of autonomous EU AFSJ concepts or systems. It also appears to indicate a reduction in international influence upon EU AFSJ law.
In general, highly disparate legislative practices appear to have evolved throughout the two AFSJ legislative cycles post-Lisbon, with a significant shift away from international law in the second legislative cycle in the context of widening and deepening of legislative packages. The paper argued that the EU actively promotes international law in its AFSJ law-making, especially human rights instruments, even in difficult and sensitive topics in which consensus is challenging. However, this is not the overall picture developed. Rising international law in directives perhaps indicates practices of legitimation and standard-setting (e.g. maximum harmonisation and innovation in law-making). On the other hand, there was subsequently a decline in international law influence in the AFSJ cycles studied here. Declining international law appears to relate to the rise of autonomous concepts − also manifested now in directives − where the EU develops sophisticated but far-reaching EU systems, actors, and programmes, mostly wholly unconnected to international law.
The EU consistently shows a tangible intent to permit the influence of international law upon the AFSJ which supports well its general efforts to participate and engage as a global legal actor even in situations where this is not required, fully legalised or regularised. Yet this is also conversely not true in places where it evolves its own autonomous idea of the AFSJ. Many recent developments in international law practice in AFSJ law-making give significant insights into the two legislative cycles taking place in these complex evolving fields as sketched here through the study of two legislative cycles and show the value of descriptive and empirical studies of the evolution of international law in EU law.