European Arrest Warrants,the rule of law and communication: What future for mutual recognition?

Abstract

This contribution addresses the insistence of the Court of Justice of the European Union on the two-step test in the context of European Arrest Warrants (EAWs) and focuses on the advantages this approach offers whilst acknowledging its downsides. Crucially, the Court's approach keeps the vertical and horizontal channels of communication open and is flanked by a subtle broadening of the criteria, which the requested judicial authority can take into consideration when assessing the second limb of the test. This shift provides national judicial authorities with tools to respond to rule of law violations by refusing the execution of an EAW. This move to judicial subsidiarity creates pitfalls, but the Court's focus on communication has the potential to transform the Court's initial top-down approach to mutual trust into a bottom-up approach, which could foster the emergence of, and strengthen, real trust between national judicial authorities.

1. Introduction

The dismantling of the rule of law poses a threat to the EU, which appears to be paralysed in the face of a hollowing out of its own values. EU institutions are criticized for insufficiently responding to the violations of core EU values, a criticism that also extends to the Court of Justice of the EU (CJEU or Court). This criticism was particularly visible in the context of the question of how to proceed with European Arrest Warrants (EAWs) issued by Polish judicial authorities given the erosion of their independence. The Polish general election on 15 October 2023 and the subsequent change in government have altered the country's direction, but the previous developments in Poland could be replicated in other EU Member States, which is why it is useful to discuss judicial responses to these threats.

This contribution takes the much-criticized L and P judgment, in which the Court upheld the two-step test, as its starting point. Whilst many have criticized the judgment, ¹ the focus of this contribution rests on discussing the opportunities that the L and P judgment and subsequent judgments offer. To this end it addresses the question of what would have happened had the Court dropped the second step of the test and the advantages of, and strategic reasons for, maintaining the second step.

Building on the maintenance of the two-step test, this contribution argues that the Court does not so much insist on mutual recognition, but that its focus rests on communication between judicial authorities. The Court's adherence to the second step of the test, which necessitates the requesting and the rested judicial authority to enter into a dialogue, provides the possibility to strengthen channels of communication, and to shape and further develop the transnational judicial network. Ultimately, the communication between judicial authorities creates a real chance for building up and further enhancing networks of cooperation between judicial authorities, which in turn allows them to foster trust.

Simultaneously, there are inherent risks in Court's approach. The first risk results from the position of requested courts, who have to assess, if there are indications to this effect, an erosion of the rule of law in the issuing Member State. This can lead to unequal power dynamics, which is why this process should be moderated to avoid and reduce frictions.

The second risk stems from the fact that the Court places the judicial responsibility for responding to rule of law (RoL) erosions on several shoulders by integrating all Member State courts in this process. This move to judicial subsidiarity, which ensures participation of the relevant players, also poses challenges, as it risks a fragmentation resulting from the different national judicial responses.

The contribution is structured as follows. First, it provides a short introduction to the principle of mutual recognition in the Area of Freedom, Security and Justice (AFSJ), and exceptions to mutual recognition in the context of the EAW, by addressing the two-pronged test (section 2), which allows the requested authority to (temporarily) refrain from executing the EAW. Section 3 addresses the situation in which the Court finds itself, and argues that both options, maintaining or dispensing with the two-step test, came at a price. To that end, the section 3.A addresses the question of what would have changed had the Court dropped the second prong of the test. Here, the focus rests on the impact of such an approach on the vertical level of cooperation, hence the communication between national courts and the CJEU (section 3.A.1), and on the horizontal level of communication between national courts (section 3.A.2). Subsequently, the focal point shifts to the opposite direction and addresses the structural and strategic reasons for maintaining the two-pronged test (section 3.B), in particular the separation of powers (section 3.B.1) and the transnational nature of the EAW (section 3.B.2). The insistence on the second step places national judicial authorities at the forefront, which is structurally in line with the transnational nature of the Framework Decision on the European Arrest Warrant ² (FDEAW). This directs the discussion to the horizontal levels of communication (section 4), as the CJEU assigns national judicial authorities a key role in the judicial response to the RoL backsliding. First, approaches of national courts to Polish EAWs are highlighted (section 4.A). Then the Court's case law, which subtly broadened the criteria, which requested judicial authorities can take into account and which affords them more room of manoeuvre to refuse EAWs from Member States with RoL breaches, (section 4.B) is discussed in more detail. This shift in Court's case law enables the requested judicial authority to conduct a more comprehensive assessment and thereby enhances the operability of the two-step test. Subsequently, the need for enhancing coherency between the approaches of national judicial authorities, and the need for a constructive communication building on, and fostering, mutual respect is highlighted (section 4.C). Moreover, it briefly discusses tools and practical support that is available to improve coherency and constructive dialogue. The final section provides concluding remarks (section 5).

2. Mutual trust, mutual recognition and the two-step test

This section briefly discusses the importance of mutual trust and mutual recognition for the functioning of the AFSJ and highlights its shortcomings with regard to human rights violations. To avoid a facilitation of human rights violations through operation of mutual recognition, the CJEU introduced the two-prong test, which in principle has to be followed, save in highly specific cases.

In 1999, the Tampere European Council Presidency Conclusions called for mutual recognition of judicial decisions and judgments in the AFSJ to be enhanced and for mutual recognition to become the cornerstone of judicial cooperation, ³ which is reflected in recital 6 of the preamble and Article 1(2) of the FDEAW. The FDEAW was the first instrument of the AFSJ to implement the principle of mutual recognition. ⁴ The notions of mutual recognition and mutual trust are closely related, ⁵ but they are distinct concepts. ⁶ The importance of both concepts and their distinctness are encapsulated in the Court's statement that

(b)oth the principle of mutual trust between the Member States and the principle of mutual recognition, which is itself based on the mutual trust (…) are, in EU law, of fundamental importance given that they allow an area without internal borders to be created and maintained. ⁷

In the AFSJ, the Court has long prescribed ‘trust’, by requiring national authorities to assume their counterpart's fundamental rights compliance. Mutual trust thus operated as a horizontal irrebuttable presumption of Member States’ fundamental rights observance. This irrebuttable presumption could not be maintained, as it facilitated fundamental rights breaches, which was exposed by the judgment of the European Court of Human Rights (ECtHR) in M.S.S. v. Belgium and Greece. ⁸ In the M.S.S. case, Greece was held to have violated Article 3 ECHR, and Belgium, which complied with EU law, and transferred the asylum seeker to Greece, which was the Member State responsible for assessing the asylum application under the (old) Dublin (II) Regulation, ⁹ was also held to be in violation of Article 3 ECHR. The M.S.S. case constitutes the turning point for the EU's AFSJ instruments that are based on mutual trust, as it demonstrated that Member States’ compliance with mutual trust does not save them from being found in violation of their ECHR obligations by the ECtHR.

The ECtHR's M.S.S. judgment necessitated a change in the EU's approach to mutual trust and prompted the CJEU to reject the conclusive presumption of Member States’ fundamental rights compliance in the context of Dublin transfers in the joined cases of N.S. and M.E. ¹⁰ This approach is also reflected in Opinion 2/13, where the Court held that mutual trust requires the presumption of other Member States’ fundamental rights compliance, but that this presumption can be rebutted in exceptional circumstances. ¹¹ The risk that Member States, who comply with their obligation of mutual trust, help facilitating another Member State's human rights violation, did not remain confined to the Dublin regime, but also features in the EAW mechanism.

A. The two-pronged test

According to the FDEAW, the executing authority can only refuse surrender if one or more of the grounds for non-execution of the European Arrest Warrant are fulfilled. Mitsilegas rightly notes that non-compliance with fundamental rights is not listed as a ground for non-execution. ¹² In order to avoid facilitating a human rights violation through operation of the FDEAW, the CJEU developed the two-step test in its 2016 landmark judgment Aranyosi and Căldăraru. ¹³ The case concerned the conditions of incarceration and their compatibility with Article 4 of the Charter of Fundamental Rights of the European Union ¹⁴ (Charter or CFR). The two-step test was subsequently also applied in the context of the erosion of the RoL, in the 2018 LM case ¹⁵ (the ‘LM test’) regarding the fundamental right to access to an independent and impartial tribunal (Article 47(2) Charter), and concerning the right to a fair trial before an independent and impartial tribunal previously established by law (X. and Y. ¹⁶ ). Moreover, the two-step test was also applied regarding the risk of being tried by a court that manifestly lacks jurisdiction (Puig Gordi and Others ¹⁷ ) and in the context of Framework Decision 2008/909 (Staatsanwaltschaft Aachen ¹⁸ ).

The two-step test requires the requested judicial authority to assess, as a first step, based on ‘objective, reliable, specific and properly updated’ ¹⁹ information, whether there is a real risk of a breach of the fundamental right in the issuing Member State, on account of systemic or generalized deficiencies (a risk in abstracto). ²⁰ In the context of the RoL, the Commission's reasoned proposal to the Council ²¹ is particularly relevant for this assessment, ²² as well as the Venice Commission's report, which features prominently in the Commission's reasoned proposal. ²³ If the executing judicial authority finds that such a risk exists, it must conduct an individualized assessment in a second step and examine whether these deficiencies could pose a real risk to the right(s) of the person concerned in the specific case. ²⁴ In order to establish the individual risk (risk in concreto), the executing authority must assess ‘in the light of the specific concerns expressed by the individual concerned and any information provided by him’ whether he ‘will run a real risk of breach of his fundamental right to an independent tribunal’. ²⁵ The CJEU clarified in Puig Gordi and Others that both steps of the assessment have to be conducted consecutively ²⁶ and cannot overlap with one another. ²⁷ The test slows down the surrender process and creates additional burdens. The answer to the question as to why judicial authorities ‘would suddenly and gladly accepted obstacles to a legal procedure that has run more or less smoothly for years’ ²⁸ lies in the risk that the executing Member State could expose the requested individual to a human rights violation in the issuing state and thereby violate its own human rights obligations, which in turn can lead to a judgment by the ECtHR to this effect.

In the process of conducting the second limb of the assessment, the executing authority is obliged, according to Article 15(2) of the Framework Decision, to ‘request from the issuing judicial authority any supplementary information that it considers necessary for assessing whether there is such a risk’. ²⁹ With regard to the RoL breaches that previously occurred in Poland, the Court held that this requires a dialogue between the issuing and the executing judicial authority in which the former may ‘provide the executing judicial authority with any objective material on any changes concerning the conditions for protecting the guarantee of judicial independence’. ³⁰ Bárd holds that it ‘is absurd to make a court engage in a discussion about its own independence’. ³¹ On the other hand, it could be argued that such a dialogue constitutes an opportunity for judges to outline the changes that resulted from authoritarian reforms and the implications that these changes had for their work. The Court's insistence on communication ensures that national judges are at least in contact with their counterparts in other Member States. This could be particularly important if they are hesitant (or afraid) to make preliminary references to the CJEU, because they fear retaliation measures from the executive. However, much will depend on the actual communication between the courts and how freely it can be conducted.

If both limbs of the test are answered positively, the executing authority must, based on Article 1(3) FDEAW, ‘refrain from giving effect to the European arrest warrant’. ³² If only the first prong has been proven, but not the second prong (the risk in concreto), or the converse situation, the surrender cannot be postponed. This approach is consistent with the Court's approach to interpret exceptions, in this case the exceptions to the constitutional principle of mutual recognition, restrictively. ³³ The prohibition to assess a risk in concreto in the absence of a risk in abstracto can be problematic for national judicial authorities in light of their obligations under the ECHR. In 2023 Krommendijk rightly notes that nothing precludes the Strasbourg Court ‘from finding an individual case when there are no structural problems’. ³⁴ Indeed, already in 2016 Hong highlighted that an individual might be subject to a serious risk of a human rights violation, even though this might be the first (documented) breach of a human rights violation in a certain place. ³⁵ In such a scenario the CJEU's assessment would end at the first prong, due to the lack of a risk in abstracto. To ignore such a case, Hong argues, could lead to conflict with the case law of the ECtHR. ³⁶ There might be cases in which the executing judicial authority is barred from assessing the second prong of the test due to absence of the first prong. Callewaert rightly asks ‘whether the two-step approach, with the individual test made contingent on the outcome of the general test, leaves enough room for the judges of the executing Member State to conform with their duties under the Convention’. ³⁷ In E.D.L. and C.K. the CJEU did assess the second step even in the absence of the first step, which might indicate a change in the CJEU's case law.

B. E.D.L. and C.K. and Others

In both the E.D.L. and the C.K. case, the Court did not follow its two-pronged test, which raises the question of whether the Court is departing from this test and whether this could be utilized as an argument in the context of the RoL erosion. Even though the discussion in the context of the RoL erosion focuses on the question of whether the assessment of the second step can be discarded, while the Court's decisions in the E.D.L. and the C.K. cases dispensed with the first step of the assessment, the latter judgments could point toward a possible change in the Court's case law.

Whereas both N.S. and M.E. (Dublin) and Aranyosi and Căldăraru (EAW) concerned the risk of a human rights violation resulting from the reception or incarceration conditions in the receiving Member State, the risk of (a) human right(s) violation can also stem from the transfer (C.K. and Others) and a fear of incarceration (E.D.L.) respectively. The case of C.K. and Others (2017), concerned a Dublin transfer and the Court held that the transfer of a person

whose state of health is particularly serious may, in itself, result, for the person concerned, in a real risk of inhuman or degrading treatment (…), irrespective of the quality of the reception and the care available in the Member State responsible for examining his application. ³⁸

Even in the absence of systemic or generalized flaws in the receiving Member State, the Court held that the person can only be transferred if it can be excluded ‘that that transfer might result in a real and proven risk of the person concerned suffering inhuman or degrading treatment’. ³⁹ Krommendijk and de Vries describe this as a trend towards a more individualistic assessment, ⁴⁰ which already features in the 2016 Ghezelbash judgment. ⁴¹

Similarly, but with regard to an EAW, the E.D.L. case was characterized by an absence of a risk in abstracto, but the person whose surrender was sought suffered ‘from serious chronic illnesses of indefinite duration’. Moreover, this condition was ‘likely to deteriorate significantly’ if the issuing Member State decided to detain him. ⁴² The CJEU referred to Article 23(4) of the FDEAW according to which surrender may be temporarily postponed for serious humanitarian reasons, and held that where surrender would expose the person ‘to a real risk of a significant reduction in his or her life expectancy or of a rapid, significant and irreversible deterioration in his or her state of health’, the executing authority must postpone surrender. ⁴³ It held that if ‘that risk cannot be ruled out within a reasonable period of time’, ⁴⁴ then the executing authority must have regard to Article 1(3) which allows this authority to refrain from surrender, ⁴⁵ which is, ‘intended to be an exception which must be interpreted strictly’. ⁴⁶ The CJEU underlined the exceptionality of non-surrender, by requiring this staggered approach whereby surrender can first be postponed provided that certain conditions are met, and, as a last resort, the authority exceptionally can refrain from surrender. In the E.D.L. case the Court transposed the approach it took in C.K. and Others (Dublin) to an EAW, and thereby provided for a parallelism between these two mutual recognition regimes in the AFSJ. In the E.D.L. judgment of April 2023, the second step of the test was discussed in the absence of the first step, which prompts the question of whether the Court departs from the two-pronged test. However, the approach in E.D.L. does not seem to apply to cases that do not share these highly specific features. This is consistent with the GN judgment of December 2023, which concerned the EAW transfer of a mother living with a child. ⁴⁷ The Court not only confirmed that the two-step test has to be followed but it also clarified that the requested judicial authority may not request information regarding the second step, in the absence of the first step. ⁴⁸ Callewaert correctly points out that the CJEU in its GN judgment not only precludes executing authorities from assessing the individual risk (second step) in the absence of systemic or generalized deficiencies (first step), but that it imposes a strict ban on an autonomous assessment of the second step. ⁴⁹

This subsection discussed cases in which the second step of the test was assessed despite the absence of the first step. Even though the discussion on EAWs in the context of the RoL erosion focuses on the converse question, namely whether the second step can be discarded, the Court's case law demonstrates a strict adherence to the two-step test, save in highly specific cases. The next section discusses the hypothetical effects of a discarding of the second step of the test, and the reasons for maintaining it.

3. The two-pronged test and beyond

This section addresses the possible effects the CJEU's dispensation with the second step of the test (section 3.A) would have had on the vertical level (section 3.A.1) and the horizontal level of communication (section 3.A.2). Moreover, this section discusses the strategic reasons for maintaining the two-step test (section 3.B) in light of the separation of powers (section 3.B.1) and the transnational nature of the EAW (section 3.B.2).

In its preliminary reference in the L and P case (Openbaar Ministerie – Independence of the issuing judicial authority), the District Court Amsterdam asked the CJEU whether it may deny the status of issuing judicial authority to the requesting Polish court in light of the ‘systemic or generalised deficiencies concerning the independence of the judiciary in the issuing Member State’ and whether it may thus presume that the second step of the test has been fulfilled. ⁵⁰ Subsequently, similar preliminary questions reached the Court. ⁵¹ The Court did not accept the invitation by the District Court Amsterdam and held that judicial authorities cannot abandon the two-step test. It stated that the existence of systemic or generalized deficiencies (first step of the test) ‘does not necessarily affect every decision that the courts of that Member State may be led to adopt in each particular case’, ⁵² which is why an individual assessment (second step) has to be conducted. Crucially, the Court held that an interpretation to the contrary, hence a reliance on the first step only, ‘would mean that no court of that Member State [Poland] could any longer be regarded as a “court or tribunal” for the purposes of the application of other provisions of EU law, in particular Article 267 TFEU’. ⁵³ Moreover, the Court clearly demarcated the limits of its interpretative margin by referring to recital 10 of the FDEAW, according to which the implementation of an EAW may only be suspended following a European Council Decision pursuant to Article 7(2) TEU. ⁵⁴

The Court's insistence on the two-pronged test has been subject to much criticism. ⁵⁵ The rejection of the possibility to suspend cooperation has been said to further legitimize changes in Poland and elsewhere. ⁵⁶ What would have been the possible effects of the Court's departure from the second step of the test?

A. Possible effects of the Court's dispensation with the second step of the LM test

Had the CJEU accepted the invitation by the Amsterdam District Court, and had it allowed national judicial authorities to dispense with the second step of the LM test, this would have suggested that the first step of the test, the existence of systemic or generalized deficiencies, is sufficient for non-execution of an EAW. In the case of Poland that would have meant that requested national judicial authorities can refrain from executing Polish EAWs, given that the first step of the test is fulfilled in light of the well-documented attacks on the independence of the Polish judiciary. This would have resulted in a suspension of the mechanism of mutual recognition of EAWs in relation to Poland, and it would have sent a clear signal that the CJEU is prepared to take practical and far-reaching measures, until the independence of the judiciary is restored. Moreover, such an approach would have sent a clear signal to other Member States, which might in the future seek to undermine the independence of their judiciaries, that they will be excluded from participation in the EAW system. Whether such an approach would be effective to change these governments’ course of action remains speculative. Suffice it to say that Grabowska-Moroz aptly points out that it ‘seems that the Member States’ governments that violate common EU values are less “concerned” about potential limitations of mutual trust than the national courts of other Member States that have to recognize and enforce the judicial decisions of such States’. ⁵⁷

A dispensation with the second step would have had far-reaching consequences for Polish courts but also for courts in other Member States. Koen Lenaerts’ speech at the Supreme Administrative Court of the Republic of Poland in March 2018 is instructive, as he drew a distinction between vertical and the horizontal levels of cooperation and highlighted the importance of judicial independence as a prerequisite for both types of cooperation.

1. Vertical level of communication and cooperation

This subsection examines the effects of a dispensation with the second step of the LM test on the vertical level of communication between national courts and the CJEU and highlights why it is undesirable to cut off Member State courts from this communication. Regarding the vertical level of cooperation, Lenaerts stated that from a supranational perspective judicial independence is a prerequisite for the preliminary reference procedure as only independent courts can be trusted with the effective and loyal application of EU law. ⁵⁸ This statement also resonates in the L and P judgment. In L and P, the Court states that the existence of systemic or generalized deficiencies ‘does not necessarily affect every decision’ by a Polish court ⁵⁹ and that therefore the individual assessment, hence the second step of the test has to be conducted. Moreover, the Court holds in L and P that an interpretation to the contrary, hence a dispensation with the second step, ‘would mean that no court of that Member State [Poland] could any longer be regarded as a “court or tribunal” for the purposes of the application of other provisions of EU law, in particular Article 267 TFEU’. ⁶⁰ Krommendijk and de Vries also refer to the importance of the second step of the test, and highlight that the insistence on the second step allows the CJEU to ‘maintain that not all courts in a particular member state are no longer independent’ ⁶¹ because the opposite finding would have implications for the application of Article 267 TFEU in relation to Polish courts. A statement by the Court to the effect that the second step of the test (individual assessment) is not necessary, implies that (all) ⁶² Polish courts have lost their independence, and would have meant that Polish courts are excluded from the preliminary reference procedure. ⁶³ Such an approach would exclude the so-called ‘last cry for help’ cases, in which Polish judges ask the CJEU whether they themselves are independent enough in terms of EU law and invite the CJEU to find that their independence or their institution's independence has been undermined. ⁶⁴ Moreover, the effects of an exclusion of Polish courts and tribunals from the Article 267 TFEU procedure would not only have consequences for (Polish) courts. If Member States’ courts fell ‘outside the notion of “court or tribunal” for the purposes of Article 267 TFEU’, this would result in a weakening of ‘the system of protection for individuals (…), hindering the effectiveness of EU law’. ⁶⁵ In addition, Reyns highlights that the exclusion of national (Polish) courts from the Article 267 TFEU mechanism has implications for EU law, as Polish courts will continue applying EU law. Without having the possibility of making references pursuant to Article 267 TFEU, Polish courts might apply EU law differently and abet the fragmentation of EU law and hamper its effectiveness. ⁶⁶ These considerations highlight why the Court did not abandon the second prong of the test and thereby a the presumption of independence. Nonetheless, in the Getin Noble Bank case the Court demarcated the limits of the presumption of independence, ⁶⁷ and in L.G. the CJEU even rebutted the presumption when it held ‘that the panel of judges of the Chamber of Extraordinary Control and Public Affairs (…) does not constitute a “court or tribunal” within the meaning of Article 267 TFEU’. ⁶⁸

Ultimately, had the Court stated that the second step of the test was no longer necessary, since Polish courts had lost their independence, it would have rendered the (previous) Polish government's restriction ⁶⁹ of the right to make references to the CJEU superfluous. In the future, it would be sufficient for a government to undermine the independence of its judiciary to cut it off from any communication with the CJEU. Therefore, the focus should rest on strengthening national courts by maintaining the dialogue with them, keeping them within the network of European courts, and supporting them in their struggle to fight off interference by the executive.

2. Horizontal level of communication and cooperation

In addition to the vertical level, the horizontal level of communication would be affected by a dispensation with the second prong of the LM test. Lenaerts stated in his Warsaw speech that from a transnational perspective, judicial independence is a precondition for national courts to trust each other, ‘so as to engage in a transnational dialogue within the framework of the AFSJ’. ⁷⁰ Had the CJEU abandoned the second step of the LM test, it would have meant that the individual assessment is no longer necessary, as requesting Polish courts have lost their independence. This would have cut off Polish courts from any communication with their sister courts within the AFSJ framework. Such a move would have had substantial implications for Polish courts and courts in other Member States. If Poland had been excluded from the EAW system, it would have been harder for Polish judicial authorities to get hold of criminal offenders prosecuted by Poland, ⁷¹ which might have entailed a high risk of impunity of the requested person. ⁷² Conversely, the exclusion of Polish courts from the EAW system could also have consequences for the other 26 Member States. A one-sided exclusion – that is, 26 Member States do not execute EAWs from Polish courts, whereas Polish courts would still be obliged to execute EAWs from these Member States – goes against the idea of reciprocity that is inherent in the concept of mutual recognition. Consequently, Poland could refuse to execute EAWs from other EU Member States. This would allow criminal offenders from other EU Member States to ‘escape’ to Poland, as it no longer participates in the EAW system and its judicial authorities can consequently no longer act as an executing authority. The exclusion of one Member State could have severe consequences for the other 26 Member States and would abet impunity of criminal offenders.

In sum, depriving courts, whose independence has been undermined, from communication with their sister courts and communication with the CJEU is likely to play in the hands of those who undermined the judiciary's independence and could have, as this subsection discussed, severe implications for the fight against impunity.

B. Strategic reasons for maintaining the second prong of the LM test

This section explores the strategic reasons for adhering to the two-stage test. The CJEU developed the two-stage test in the context of Article 4 CFR in the Aranyosi and Căldăraru judgment, and subsequently applied it to the right to a fair trial (Article 47(2) CFR) in its LM judgment. Wendel convincingly argues that the CJEU could hardly impose less demanding requirements for refusing the execution of an EAW when the relative right of Article 47(2) CFR is at stake, than when the absolute right of Article 4 CFR is invoked. ⁷³ He adds that the fundamental rights framing of the RoL erosion prevented the Court from dropping the second limb of the test in the context of Article 47(2) CFR without risking an inconsistency with its case law in the context of Article 4 CFR. ⁷⁴ Indeed, the Court could hardly discard the second limb of the test in the context of relative right of Article 47(2) CFR, thereby making it easier to refrain from executing the EAW in the context of the RoL erosion, whilst maintaining the two-stage test, which imposes a higher threshold for non-execution of the EAW, in the context of the absolute right of Article 4 CFR. Next to this convincing argument, the following subsections highlight why the departure from the second prong of the LM test would have been difficult to reconcile with the separation of powers (section 3.B.1) and with the structure of the FDEAW (section 3.B.2).

1. Separation of powers

In light of the subversion of the separation of powers that previously occurred in Poland, this subsection sets out why it is important that the Court refrains from encroaching on the powers of the European Council or the legislator.

Frackowiak-Adamska correctly points out that literal interpretation is not the CJEU's favourite tool, as the Court has often developed its case law by reference to the spirit of the Treaties. ⁷⁵ Therefore, she poses the question as to why the Court put so much emphasis on recital 10 of the preamble to the FDEAW. ⁷⁶ In the L and P judgment the Court referred to recital 10, ⁷⁷ which makes the suspension of the EAW mechanism dependent on a European Council Decision pursuant to Article 7(2) TEU. The Court's reference to recital 10, could be seen as an indication that the Court does not wish to interfere with the competences of the European Council, ⁷⁸ whose exclusive prerogative it is to suspend the EAW in relation to a Member State, ⁷⁹ or with the competence of the legislator. Even though the recitals of the preamble are not legally binding, they express the intention of the legislator. The Court's self-restraint does indeed seem to contradict its approach in earlier judgments when it acted as the motor of integration by creatively interpreting the Treaties. Moreover, it was the CJEU who established the two-step test, which is not contained in the FDEAW. However, the Court established the test, based on Article 1(3) FDEAW, to avoid that a ‘blind’ application of mutual recognition leads to a human rights violation. The two-step test constitutes a high threshold that ensures that non-execution of the EAW remains the exception. Dispensing with the second step of the test, would have lowered the threshold considerably and would have led to a suspension of EAWs in relation to Poland.

A suspension of EAWs in relation to Poland would have meant a departure from the wording of the FDEAW including its preamble, which could have invited criticism that the CJEU itself does not do what it preaches. If the Court assumed competences that were not assigned to it and suspended the EAW mechanism, the issue of the Court's compliance with Treaty obligations could be raised, ⁸⁰ which could ultimately undermine its own legitimacy and credibility. The previous Polish government subverted the separation of powers by interfering with, and curtailing, the competence and independence of the judiciary. The CJEU is therefore particularly cautious to stay within the interpretative limits of the FDEAW to avoid any criticism that it is acting outside its competences. This subsection explained why it is important that the Court adheres to the requirements set out in the preamble, especially against the background of the violation of the separation of powers that previously occurred in Poland.

2. The transnational nature of the EAW system

The FDEAW builds on mutual trust and is designed as a horizontal mechanism, which facilitates communication and cooperation between judicial authorities in different Member States. It aims at ensuring a transnational reach of national decisions to facilitate the swift surrender of suspected or convicted persons. If the CJEU had dispensed with the second prong of the test, it would have prescribed distrust and forestalled transnational communication, which is difficult to reconcile with the purpose and structure of the FDEAW.

First, a dispensation with the second step of the test would have led to a ‘blind’ non-execution of EAWs. A blind non-execution is as problematic as a blind execution of EAWs based on blind trust, as both scenarios do not account for the specific situation of each individual case. Moreover, a blind non-execution would have fostered blind distrust, as it would have sent the signal to the executing judicial authorities that Polish judicial authorities can no longer be trusted, a signal which is to be avoided.

Second, a dispensation with the second step of the test would have pre-empted communication. The two-pronged test requires a communication between the requesting and the requested judicial authority and makes the execution of the EAW dependent on the outcome of the latter's assessment, which bases, among other factors, on the information provided by the former. Every requested judicial authority must determine, on a case-by-case basis, whether the accused or convicted person can be surrendered. The FDEAW distinguishes between EAWs for the purpose of prosecution and EAWs for the surrender of a convicted person. ⁸¹ Frackowiak-Adamska highlights that the former ‘differ from classic recognition of judgments, as they concern surrendering a person to the criminal justice system of another State, with future events which are difficult to evaluate’. ⁸² Regarding situations in which the requested person has already been convicted, the verification of a breach is easier. This approach also seems to be reflected in the decision of the District Court Amsterdam in the L and P case. P was convicted in 2018 on various counts of threatening behaviour and ill-treatment, and the District Court Amsterdam executed the EAW for his surrender. The District Court did not find any reasons to assume that the fundamental right to an independent court had been violated and held that it had no reason to assume that no fair trial has been conducted in 2018. ⁸³

The EAW regarding L, by contrast, was issued for the purposes of conducting a criminal prosecution against him. L was charged with drug trafficking and possession of false identity documents. The District Court Amsterdam assessed the facts of the individual case and decided not to execute the Polish EAW against him. ⁸⁴ It must also be noted that the facts of L's case were very specific. The District Court stated that L's name and information relating to his case were mentioned in a general memo from the Public Prosecutor’s Office in Poland, ⁸⁵ and his case also attracted the attention of both the media and politicians in Poland. ⁸⁶ According to the District Court, L was therefore subject to the special attention of the authorities, which meant that there was a risk that the ‘chilling effect’ (of disciplinary proceedings against judges) would have a concrete impact on his procedure. ⁸⁷ Moreover, it noted that in response to its referral of L's case to the CJEU, judicial officers in Poland had been ordered to assess, regarding EAWs issued by Dutch authorities, whether mandatory grounds for refusal apply. ⁸⁸ Even though the District Court's decision to execute the EAW regarding P and not to execute the EAW in L's case demonstrates that the application of the LM test at the national level facilitates a more differentiated assessment, it must be highlighted that the situation in L's case is unique. ⁸⁹ In fact the Dutch ImprovEAW report of October 2021 refers to the L case as the only case in which an individual risk was established. ⁹⁰ The report states that between 17 December 2020 and 1 May 2021, a general risk was established in 21 final decisions on Polish execution- and prosecution-EAWs, but that surrender was not blocked because an individual risk had not been established. ⁹¹ In L's case, by contrast, the heightened interest in the case from both Polish politics and media, was an important factor in the national court's reasoning. Such a level of attention is not necessarily present in every case and therefore the effects of this case should not be overstated. ⁹²

This subsection has highlighted that communication between national judicial authorities, which is envisaged by the second limb of the test, is system-compliant and allows a differentiation depending on the facts of the individual case. Nonetheless, the system can pose considerable challenges for the executing judicial authority, depending on whether they receive answers from the issuing authority, the quality of these answers and their reliability.

4. Horizontal levels of communication and the role of national courts

The CJEU assigns national courts a core role in responding to RoL breaches, for which it places a strong focus on inter-court communication. The role of national courts and their communication shall be the focus of this section. The Court's insistence on the second prong of the LM test distributes the responsibility for assessing and addressing the erosion of the RoL on several judicial shoulders. But it is also said that this task amounts to a Herculean task, ⁹³ as the individual concern is extremely difficult to prove, ⁹⁴ making the second limb of the test unworkable ⁹⁵ and unfeasible. ⁹⁶ Martufi and Gigengack, who analyse the application of the two-stage test by the District Court of Amsterdam in decisions issued between June 2016 and June 2020, point out that even though the persons requested for surrender argued in numerous cases that an individual risk exists, the District Court has not been able to find sufficient information to end the surrender procedure. ⁹⁷ This finding is also reflected in the above-mentioned Dutch ImproEAW report. ⁹⁸ However, Mancano rightly points out that the threshold of the second prong of the test might not be as insurmountable as it seems, since the Court leaves the requested judicial authority some degree of flexibility. ⁹⁹ Indeed, the CJEU subtly modified the parameters of the assessment of the second limb of the test in cases following the L and P judgment, which broadened the margin of manoeuvre for national courts to reject the execution of the EAW. Subsequently, some approaches of national courts to EAWs originating in Poland are outlined (section 4.A) before the subtle shift in the CJEU's case law (section 4.B) is discussed. Then the need for improvement of the coherency of national approaches and the need for fostering mutual trust through constructive communication (section 4.C) is addressed.

A. National judicial authorities’ approaches to Polish EAWs

First, it must be noted that not all Member States’ judicial authorities identify problems regarding EAWs issued by their Polish counterparts. The ImprovEAW Project has conducted a study of seven EU Member States, namely Belgium, Greece, Hungary, Ireland, the Netherlands, Poland and Romania. The report was based on a questionnaire to detect common issues in EAWs with the aim to propose practical tools and best practices. ¹⁰⁰ Question 52 of that questionnaire asks whether the executing authorities were confronted with a case in which they established systemic or generalized deficiencies in the judicial system liable to affect the independence of the judiciary, leading to a real risk of violation of the right to an independent tribunal. In the Belgian report, one case from 2021 was identified in relation to an EAW issued by a Polish authority, but the second step of the test was said to not to be fulfilled, with the consequence that there was no bar to surrender. ¹⁰¹ The Greek, ¹⁰² Hungarian ¹⁰³ and Romanian ¹⁰⁴ reports do not note any such cases, apart from one case concerning a Maltese request addressed to Greece. The Irish report lists several cases in relation to Poland, ¹⁰⁵ but also notes that no case has succeeded in a sense that the individual concerned would not be transferred. ¹⁰⁶ The report of the Netherlands notes several cases in relation to Poland. ¹⁰⁷ Interestingly the Polish report notes that requested judicial authorities in Poland refused to execute two EAWs from the Netherlands, citing the risks to human rights protection resulting from the lack of impartiality of the Dutch issuing authority. ¹⁰⁸

Second, even before the Court modified the parameters of the second step of the assessment, domestic courts developed approaches to reject the execution of EAWs originating in Poland. The Higher Regional Court of Karlsruhe in Germany rendered a landmark decision in 2020. ¹⁰⁹ The applicant in that case had objected his surrender to Poland as the systemic deficiencies were likely to affect the independence of the judiciary and thereby his right to a fair trial. ¹¹⁰ After addressing extensively the judicial reforms in Poland, the Karlsruhe court turned to the second step of the LM test. Since the individual risk could not be excluded, the court sent a list of questions to the Polish Ministry of Justice. ¹¹¹ The German court did not await the answers of the Polish Ministry of Justice, ¹¹² which in itself is problematic. Instead, the court set aside the extradition warrant due to a high probability that surrender would be inadmissible at that point in time. ¹¹³ Regarding the individual risk, the Karlsruhe court held that there are factual indications that a risk of a violation of the requested person's right to a fair trial, if he was surrendered, cannot be ruled out, ¹¹⁴ which shows that an individual risk can be ascertained.

The second step of the LM test, which has been described as unworkable and extremely difficult to prove, ¹¹⁵ has led to calls to lower the threshold or to adapt the second step of the LM test. Frackowiak-Adamska, who criticizes the CJEU for not being willing to drop the second limb of the LM test, argues that the adaptation of the second step seems indispensable. ¹¹⁶ She highlights the approach by the Vestfold District Court in Norway, ¹¹⁷ which did not completely dispense with the two-pronged test. Instead, as Holmøyvik points out, the Vestfold court amended the test so that ‘the greater the general risk for a breach, the less specific grounds for a breach of Article 6 of the ECHR should be required in the specific case’. ¹¹⁸ This approach might be a viable option to respond to, and accommodate, the difficulties national courts face when addressing the second step of the test. Moreover, the CJEU has broadened the room of manoeuvre for national judicial authorities.

B. The shift in the Court's case law

Even though the CJEU has not abandoned the second step of the assessment, it has taken steps to enhance the room of manoeuvre of national judicial authorities when conducting the two-pronged test. Already in the LM judgment, the CJEU is said to have granted national judicial authorities a significant discretion, and a great flexibility, especially in the context of highly politicized cases. ¹¹⁹ In the L and P judgment, the Court, whilst insisting on the two-pronged test, emphasized that the increase in systemic deficiencies must prompt the requested authority to exercise vigilance. ¹²⁰ After L and P, the Court went a step further and broadened the factors that can be taken into consideration by the requested judicial authority when conducting the individual assessment. ¹²¹ Whilst outwardly insisting on the two-step test, the modification of parameters of the second step ¹²² provides national judicial authorities with more discretion when deciding on the execution of an EAW.

In the X and Y case and the WO and JL case, the Court recalled that the second step of the LM test requires an overall assessment that must have regard to several factors. The person subject to surrender can rely ‘on any ad hoc factor, specific to the case in question (…) capable of establishing’ that the surrender ‘will tangibly undermine his or her fundamental right to a fair trial’. ¹²³ The Court highlighted that the person concerned should provide ‘information relating to, inter alia, the procedure for the appointment of the judge or judges concerned and their possible secondment to that panel of judges’ ¹²⁴ so that the requested authority can assess whether generalized or systemic risks are likely to materialize in the person concerned.

A further factor is the availability of legal remedies, in particular the possibility to request the recusal of (a) judge(s). ¹²⁵ Importantly, the Court held that any evidence that is available to the requested judicial authority ‘enabling it to conclude, as part of its sovereign discretion, that the recusal procedure or the legal remedies (…) are ineffective, will be relevant’ for evaluating the existence of a real risk of a breach. ¹²⁶

Interestingly, the Court also clarified that the overall assessment of the circumstances is based on the ‘information provided by that person and supplemented, where appropriate, by the information provided by the issuing judicial authority’. ¹²⁷ The choice of words demonstrates that the question of whether and to what extent it is appropriate to consider the material provided by the issuing authority (which might have been ‘captured’), will be answered by the requested authority. The Court grants the national judicial authority discretion not to consider the information at all, should it find it inappropriate to consider the material provided by the requesting authority.

Moreover, any conduct by the issuing authority showing a lack of sincere cooperation

may be regarded by the executing judicial authority as a relevant factor for the purposes of assessing whether the person (…) runs a real risk of breach of his or her right to a fair trial before a tribunal. ¹²⁸

The breadth of the national court's margin to consider a broad range of material from a variety of sources is demonstrated by the Court's statement that the requested authority can consider ‘all the information which it considers to be relevant’. ¹²⁹ Concerning the source of this information, the CJEU states that ‘that information may, in particular, relate to statements made by public authorities which could have an effect on the specific case in question’. ¹³⁰ The choice of words ‘in particular’ clarifies that other statements and information can be considered. Regarding the probability of these factors’ impact on the individual case, the Court does not impose a high threshold. It refers to statements ‘which could have an effect’. Mancano points out that the specific material that is considered by the requested authority does not have ‘to prove untrustworthiness and risk of violation with absolute certainty’. ¹³¹ The exact level of certainty that the risk materializes seems to remain subject to the margin of discretion of the requested judicial authority.

Moreover, the requested authority may rely on any other information ‘such as information relating to the personal situation of the person concerned, the nature of the offence for which that person is prosecuted and the factual context in which the European arrest warrant concerned is issued’. ¹³² As indicated by the notion ‘such as’, the list of criteria is inexhaustive.

This underpins the importance of judicial authorities’ access to updated and accurate information which is, for example, available through the European Judicial Network's (EJN) Judicial Library. ¹³³ The order in WO and JL as well as the judgment in X and Y provide national courts with an enhanced discretion concerning the factors that they can take into consideration when assessing the individual risk. Moreover, it seems that this discretion not only relates to factors that can be taken into consideration, but also the weight that the requested court can attach to these factors in its balancing process. The Court strengthens the role national judicial authorities in responding to the RoL erosion, coupled with broadening their discretion. At the same time this move to judicial subsidiarity increases the risk of fragmentation resulting from different judicial approaches. ¹³⁴ This risk could be countered by an exchange of information and a streamlining of processes through the EJN and Eurojust, which will be addressed in the next subsection.

C. Enhancing coherency and fostering mutual trust through constructive communication

The Court places the judicial responsibility for responding to the RoL erosion on several shoulders by integrating all Member State courts in this process. This move to ‘judicial subsidiarity’ enhances the risk of different judicial authorities responding differently. For an informed response, national judicial authorities must have a considerable expertise and background knowledge of developments in countries with a RoL erosion to adequately conduct their assessment and to respond to the surrender request. Therefore, an exchange of information between national courts, in which they share their knowledge and approaches, is crucial. In the Netherlands, EAW requests are pooled at the District Court in Amsterdam, in order to streamline the decision-making process and to avoid divergence. ¹³⁵ The District Court in Amsterdam consequently has a considerable expertise. Other countries might not necessarily operate such a system, but opted for a decentralized one with the consequence that at some point a national court might face its first EAW request. In such a case, a transnational exchange of information and expertise would be useful and could be coordinated by Eurojust, ¹³⁶ which is also mentioned in the FDEAW. ¹³⁷ The 2020 European Implementation Assessment for the European Arrest Warrant states that ‘Eurojust plays a crucial role in facilitating judicial cooperation’ and highlights that there ‘is still untapped potential of making use of this EU agency's service’. ¹³⁸ The Implementation Assessment identifies ‘(t)raining and exchanges of judicial authorities’ as an area which ‘could be further promoted and funded’, ¹³⁹ and which could also be utilized in the context of EAWs. Several tools, such as the Judicial Atlas, ¹⁴⁰ Fiches Belges, ¹⁴¹ the Compendium ¹⁴² and the Judicial Library ¹⁴³ , are already available to facilitate and improve judicial cooperation. In addition, Eurojust and the EJN could develop further tools to assist decision-makers. They could foster and streamline the exchange of information and best practices between judicial authorities, and thereby help to reduce divergences and mitigate fragmentation between national approaches.

The EJN and Eurojust could also play a major role in supporting the constructive communication between the requesting and the executing judicial authorities as equal partners. Several authors have pointed out that the CJEU's case law has turned the concept of a dialogical cooperation between equal partners into a situation where the executing authority supervises the requesting authority. ¹⁴⁴ A situation in which the issuing judicial authority feels assessed creates uneven power dynamics, ¹⁴⁵ which in turn damages trust. Needless to say, that such a situation is not conducive to creating a basis for constructive communication and cooperation, let alone for creating and maintaining mutual trust. First and foremost, the dialogue between the issuing and the executing judicial authority pursuant to Article 15(2) and (3) FDEAW must be informed by the duty of sincere cooperation (Article 4(3) TEU). ¹⁴⁶ The CJEU clarified in the X. and Y. case that a violation of the duty of sincere cooperation can work to the detriment of the issuing authority, ¹⁴⁷ which was already addressed in the previous subsection. Conversely, in ML a list of 78 questions was sent by the German executing judicial authority to the Hungarian issuing authority. ¹⁴⁸ The CJEU held that a number of questions, taken individually, were relevant for the assessment, but that ‘because of their number, their scope (…) and their content’ these questions made it impossible for the issuing authority to provide a useful answer. ¹⁴⁹ Peristeridou also points out that the unstandardized process for requesting supplementary information from the issuing authority, leads to irrelevant questions being asked, and that questions sometimes arrive on a one-by-one basis, which is far from fostering a smooth communication process. ¹⁵⁰ This point is underscored not only by the ML judgment, but by a request of the Dutch judicial authority to their Polish counterparts to provide supplementary information in the context of the rule of law erosion in Poland. In its request the Dutch court asked, according to the Polish judicial authority, ‘several irrelevant questions concerning salary of judges’ which, in the opinion of Polish judges, demonstrates that the Dutch court lacks impartiality. ¹⁵¹ The questions of the Dutch court had implications for the cooperation with their Polish counterparts, as the latter refused to execute two EAWs from the Netherlands due to concerns about the impartiality of the requesting Dutch authority. ¹⁵² Indeed, a more standardized process with clearer guidance as to what can and should be asked avoids the perception, whether justified or unjustified, that the counterpart in another Member State is asking irrelevant or biased questions.

Moreover, the Polish ImprovEAW report also notes that questions by the Dutch court to the competent Polish judicial authority were not translated into Polish. ¹⁵³ The Fiches Belges of the EJN explicitly state with regard to the ‘European Arrest Warrant (C.1)’ in relation to Poland that requests and decisions should be translated into the Polish language. ¹⁵⁴ Translating requests for additional information into the working language of one's counterpart can help speeding up the process and establishing a good work relationship, especially if the counterpart explicitly asks for this. The cooperation between national judicial authorities is still evolving, and the above-mentioned tools can also be utilized here to support and improve this process. Regarding practical support, the Joint Paper by Eurojust and the EJN of April 2024 states that the EJN should be used, among other instances, to facilitate judicial cooperation, in particular the ‘exchange supplementary information for the proper execution of the request’. ¹⁵⁵

Questions for supplementary information to the issuing judicial authority should be guided by the objective to avoid a human rights violations, and should be limited to what is necessary to exclude this risk. This objective is not always easy to implement and the type of supplementary information will also be shaped by the facts of the individual case, but several tools and practical assistance are available to judicial authorities to navigate these challenges. Above all, judicial authorities need to comply with their duty of sincere cooperation. This subsection highlighted that existing tools and available support could be utilized where needed to reduce fragmentation resulting from different judicial responses and to enhance a constructive communication of equal partners.

5. Conclusions

The Court was in a difficult position in relation to Poland, a situation that could re-emerge if another EU Member State is undermining the independence of its judiciary. In that situation, the Court could have dispensed with the second step of the LM test, which would have implied that Polish judicial authorities had lost their independence and which would have led to a suspension of EAWs in relation to Poland. Such an approach would have sent a clear message to the Polish government. The price of such a clear statement would have been that Polish courts were cut off from both vertical and horizontal channels of communication, as they would no longer qualify as actors in this communication process, due to their lack of independence.

Instead, the CJEU opted for maintaining the second prong of the test, which came at a price. The Court's silence has been criticized for further legitimizing changes in the Polish judiciary. On the other hand, its approach has kept the vertical channel of communication between itself and Polish courts open, with Polish courts still being able to make preliminary references. Had the CJEU cut off Polish courts from the preliminary reference procedure, it would have played in the hands of the previous Polish government. Second, and perhaps even more important, given the nature of the EAW system, the Court's approach allows for the horizontal communication between judicial authorities to take place. By insisting on the second prong of the LM test, the executing authority must establish, based on its assessment of the individual case, whether an EAW can be executed. National judicial authorities are the key actors in the EAW system and therefore the Court's approach, to leave the final decision to the national judicial authority, is system-compliant. By insisting on the second step of the assessment, the CJEU passes the ball back to national courts. This approach can be seen as enabling national courts to play a greater role in responding to the rule of law crisis. ¹⁵⁶ However, the CJEU's reliance on, and referral back to, national courts has been criticized as faint-hearted, linked with a call for the CJEU to take greater responsibility and refrain from delegating politically delicate tasks to national courts. ¹⁵⁷ Judicial self-restraint on part of the CJEU, or, negatively phrased, the avoidance of a clear stance by the CJEU, could perhaps be interpreted as letting down national courts, ¹⁵⁸ but it could also be seen as empowering national courts. The CJEU's approach leaves national judicial authorities a considerable margin to shape the transnational system of judicial cooperation. National judicial authorities constitute the backbone of the inter-state judicial cooperation, ¹⁵⁹ and as such they have to communicate with one another. This communication in turn is necessary to consolidate and refine the judicial dialogical network. ¹⁶⁰ Through each decision-making process that precedes the refusal or execution of the EAW, a domestic judicial authority can take part in the discussion, contribute to the development of the discourse, and shape the transnational judicial network. The CJEU's approach is not without risks. Its move to judicial subsidiarity can lead to different approaches by national judicial authorities, which underpins the need for exchange of information between judicial authorities, recourse to the available tools and the practical support that is available to enhance convergence. Moreover, the structure of the LM test abets a power imbalance, through which the issuing authority might feel mistrusted and assessed by the executing authority. The need to foster a dialogue between equal partners is key and should be supported by having recourse to the practical assistance that is provided by the EJN and Eurojust.

In judgments following the L and P case, the CJEU has broadened the factors that national courts can take into account in the individual assessment and has accorded national courts a considerable discretion in weighting these factors. The Court's insistence on the second prong of the LM test is not synonymous with an insistence on mutual recognition of EAWs. It rather obliges national courts to communicate, to exchange information and to assist each other. If the inter-court communication leads to the conclusion that an EAW request cannot be executed, both authorities have to discuss alternative means of cooperation. Thus the Court's approach can be useful, as national (judicial) authorities are not only the prime actors in the EAW system, but they might also be required to cooperate in the transfer of criminal proceedings or the transfer of sentences. These types of cooperation require, in the first place, that (judicial) authorities enter into a dialogue. In this regard the Court's L and P judgment can also be understood as a call to national courts to strengthen their communication patterns. Finally, the CJEU's approach could foster a bottom-up approach to mutual recognition. For a long time, the Court applied a top-down approach, which obliged national judicial authorities to trust each other and to execute each other's decisions. By obliging national courts to conduct an individual assessment, provided that there is a risk of a human rights violation, and making the execution of the EAW dependent on the outcome of this assessment, mutual recognition now seems to take more of a bottom-up shape. For the bottom-up approach to work, a situation where the issuing authority feels controlled is to be avoided. The principle of sincere cooperation should guide the exchange of information to foster transnational communication between national authorities. Based on the information national judicial authorities receive from their counterparts, they can decide what type of cooperation is best suited to avoid impunity of the criminal offender and to comply with their obligations under both EU and ECHR law.