Monday, 30 May 2016

Money laundering, customer due diligence and data protection: the CJEU's judgment in Safe Interenvios

Marcin Kotula, Legal Officer at the European Commission

The views expressed are purely those of the author and may not in any circumstances be regarded as stating an official position of the European Commission


The recent judgment of the CJEU in the case of Safe Interenvios was triggered by a preliminary reference from the Provincial Court in Barcelona (Audiencia Provincial). The Court in Barcelona submitted to the CJEU a number of questions related to the interpretation of the third Anti-Money Laundering Directive 2005/60/EC (AML Directive, since replaced by the fourth money laundering Directive, discussed here).

In the case in question, Safe, a company which falls under the definition of a "financial institution" within the meaning of the AML Directive and of a "payment institution" within the meaning of the Payment Services Directive 2007/64 (PSD) has been transferring the funds of its customers abroad through the accounts it held with 3 banks, BBVA, Sabadell and Liberbank. The transfers were to be carried out by agents who were accordingly authorised by Safe and verified by the Bank of Spain (Banco de España). After discovering irregularities regarding Safe's agents the banks, acting under Spanish Law 10/2010 on the prevention of money laundering and terrorist financing[1] which transposes the AML Directive in Spain requested various information from Safe. When Safe did not provide them with the requested information the banks closed its accounts. Before closing Safe's account BBVA informed SEPBLAC[2] that Safe might be involved in money laundering activities.

Safe challenged the closure of its accounts before the Commercial Court in Barcelona (Juzgado de lo Mercantil) arguing that the banks have also been transferring funds abroad and that insofar they have been competing with Safe on the same market. In consequence, according to Safe, the closure of accounts was an act of unfair competition. Safe argued further that the information requested by the banks which related to Safe's customers as well as to the origin and destination of the funds could not have been provided without breaching the data protection legislation.

Safe's challenge was largely unsuccessful as the Commercial Court in Barcelona did not find a specific infringement of competition law by none of the banks. It concluded that BBVA closed Safe's account on the basis of checks which showed that nearly a quarter of transactions were not carried out by agents authorised by Safe and verified by the Bank of Spain. As for the closure of accounts by Sabadell and Liberbank, the court in Barcelona partly ruled in Safe's favour concluding that these two banks failed to properly set out the reasons for their closures.

Subsequently Safe, Sabadell and Liberbank appealed against that judgment to the Provincial Court in Barcelona which submitted the preliminary questions to the CJEU. 

The CJEU was asked, first, whether customer due diligence measures, laid down in the AML Directive to respond to the risks of money laundering and terrorist financing, could be applied by a credit institution (in the case at hand, a bank) to a financial/payment institution such as Safe, given that financial/payment institutions are already subject to supervision by competent authorities under the PSD and the AML Directive. The CJEU was then additionally asked what type of customer due diligence measures (standard, simplified or enhanced) could be applied in such a scenario and which circumstances could trigger the application of those measures. Finally, the national court asked if the measures and the provision of certain information requested by the banks from Safe are in line with EU competition law (Safe claimed that the banks compete with it on the payment services market) and with EU data protection law (according to Safe, the banks requested the identification data of its customers and of the recipients of the funds which Safe transferred).

The AML Directive sets out the legal framework for measures aimed at preventing and combatting money laundering and terrorist financing. Its provisions are to a great extent inspired by the recommendations of the Financial Action Task Force (FATF), the main international body in the area of combatting money laundering and terrorist financing.  Article 3 of the AML Directive defines which institutions and professions are to apply the anti-money laundering measures. The list in Article 3 includes credit institutions such as banks and financial institutions such as Safe. Chapter II of the AML Directive, which deals with customer due diligence, distinguishes between 3 types of such diligence, i.e. simplified, standard and enhanced.

As far as standard due diligence is concerned, Articles 7 and 8 of the AML Directive describe in which circumstances due diligence should be applied and what measures this might involve. The latter provision underlines that the extent the due diligence measures can be determined on a risk-sensitive basis depending on the type of customer, business relationship, product or transaction.

Article 9 of the AML Directive specifies the checks that need to be undertaken before the establishment of a business relationship or the carrying-out of a transaction. It also indicates when a business relationship must be terminated or a transaction cannot be carried out.

Article 11 sets out the simplified customer due diligence measures which inter alia apply in situations where the customers are credit institutions or financial institutions. Such customers are already covered by the scope of Article 2 of the AML Directive and need to apply due diligence measures towards their own customers. Enhanced customer due diligence is dealt with in Article 13.

Pursuant to Article 37 of the AML Directive the compliance with the requirements of the Directive by the institutions and persons that need to apply it is supervised by competent authorities. Credit institutions and payment institutions are also covered by the PSD.

Payment institutions get authorised to provide payment services by competent authorities designated by the Member States. These authorities are also empowered to supervise the compliance with the requirements that are applicable to payment service providers.

The CJEU's analysis

The CJEU first dealt with the question if financial institutions such as Safe can be the addressees of standard or enhanced customer due diligence measures despite the derogation in Article 11 of the AML Directive which foresees the application of simplified due diligence measures towards financial institutions.

The Court underlined that Article 11 of the AML Directive does not derogate from Article 7(c) under which standard customer diligence measures must be applied when there is a suspicion of money laundering or terrorist financing. Thus, a national provision which authorises the application of standard due diligence measures vis-à-vis financial institutions in such circumstances of suspicion is compatible with the Directive.

In a similar vein, Article 11 of the AML Directive does not derogate from Article 13 thereof. The latter requires enhanced customer due diligence measures to be applied in situations where the risk of money laundering or terrorist financing is higher. Paragraphs (2) to (4) of Article 13 contain a non-exhaustive list of such situations which by their nature present a higher risk. Whilst this list does not include the transfer of funds abroad the Member States have a margin of discretion in applying a risk-based approach and identifying other situations which are, by their nature, associated with a greater risk of money laundering or terrorist financing. In the case at hand, the transfer of funds abroad was included by the Spanish legislator in Law 10/2010 (Article 11) as one of the higher-risk situations which trigger enhanced customer due diligence.

The CJEU then dealt with Article 9 of Spanish Law 10/2010 which on the one hand allows the non-application of standard customer due diligence towards financial institutions but on the other hand empowers the Minister of Economic Affairs and Finance to exclude the application of simplified due diligence towards certain customers. On this point, the CJEU noted that the AML Directive only lays down the minimum level of EU harmonisation with Article 5 of the Directive envisaging the possibility of adopting or retaining in force stricter provisions in the EU Member States. This conclusion was supported by an earlier CJEU judgment in Jyske Bank Gibraltar. The stricter provisions which can apply in the Member States need to serve the purpose of strengthening the fight against money laundering and terrorist financing. They may thus also relate to additional situations which, according to the Member State, present a risk of money laundering or terrorist financing  even if the AML Directive does not prescribe any type of customer due diligence for those situations.

The second group of questions before the CJEU related to the extent of powers which credit institutions may exercise in the context of customer due diligence and to the relation between those powers and the powers of the supervisory authorities under Article 37 of the AML Directive and under Article 21 of the PSD. Here, the Court noted that an institution covered by the AML Directive cannot establish a business relationship or carry out a transaction through its account or must terminate an existing business relationship when it is unable to obtain the various items of information that are defined  in Article 8 of the Directive. These items include the verification of the customer's and the beneficial owner's identity (in the latter case pursuant to a risk-based approach) as well as the information on the purpose and intended nature of the business relationship. The inability of the institution to obtain these types of information might be due to the customers' refusal to cooperate (as in the case at hand) or to other factors.

The CJEU went on to identify the limitations that need to be applied when taking a measure such as the termination of a business relationship or the refusal to carry out a transaction through the bank account. The measure must be proportionate to the risk of money laundering or terrorist financing and thus cannot be taken in the absence of sufficient information which point out to that risk.

The Court then stated that the powers exercised in the context of customer due diligence and the supervisory powers of the competent authorities under the AML Directive and the PSD are rather to be seen as separate and complementary. In consequence, a credit institution may take account of the due diligence measures which its customer had to apply towards its own customers but the extent of the credit institution's due diligence measures in such a scenario must be appropriate to the risk of money laundering and terrorist financing. In addition, a credit institution must in that case neither compromise the supervisory tasks of the competent institutions under the PSD nor replace those supervisory authorities.

Next, the CJEU spelled out the conditions in which the national legislation can authorise or require standard or enhanced customer due diligence measures towards a financial institution. The CJEU's reply to the first group of questions indicated already that such measures can be applied vis-à-vis financial institutions pursuant to Article 13 of the AML Directive (enhanced due diligence) and Article 5 (stricter provisions). In this part of the judgment however the Court examined how the Member States (when prescribing such measures) or the credit institutions (when authorised by the Member State to apply such measures) can exercise the powers under Article 5 and 13.

The CJEU started by recalling its case-law on the freedom to provide services and on the permissible restrictions of that freedom (Art. 56 TFEU). It reminded that in Jyske Bank Gibraltar the prevention of and fight against money laundering and terrorist financing was recognised as a legitimate public interest objective which could justify a barrier to the freedom to provide services. It then turned to the question if Article 11 of Spanish Law 10/2010 which identifies the transfer of money abroad as a situation which always presents a higher risk of money laundering and terrorist financing (and in consequence triggers enhanced customer due diligence) is appropriate for attaining this legitimate public interest objective. In that regard, the Court stressed that both the national legislator (when prescribing standard or enhanced due diligence measures towards a financial institution) and the credit institutions (when authorised by the Member State to apply such measures) must carry out a complete risk assessment prior to deciding on the measures to take. Such measures must furthermore be proportionate to the risk so identified. The final element of this part of the CJEU's judgment was thus dedicated to the proportionality of Article 11 of Spanish Law 10/2010. Here, the Court concluded that the restriction of the freedom to provide services laid down in Article 11 would be proportionate if no less restrictive means were available and if the restriction was compatible with the fundamental rights and freedoms under the Treaties and the Charter e.g. with the right to protection of personal data (Article 8 of the Charter) and with the principle of free competition. Whilst, in principle, leaving the protection of personal data aspects for the last part of the judgment the Court found that a less restrictive measure was available in this case. In the case at hand the Spanish legislator generally presumed that all transfers of money abroad present a higher risk of money laundering and terrorist financing whereas it could have provided a possibility of rebutting that presumption in individual cases which objectively do not present such a risk.

The last group of preliminary questions put before the CJEU focussed on the compatibility of the enhanced due diligence measures with the EU data protection law, as set out in the Data Protection Directive (Directive 95/46). The Provincial Court in Barcelona asked if Safe can be obliged to provide the banks with the identification data of its customers and in particular those from whom the transferred funds originated as well as with the identification data of the recipients of the funds. In the reply to the previous group of questions the CJEU has already indicated that the due diligence measures taken pursuant to Articles 5 and 13 of the AML Directive need to be compatible with Article 8 of the Charter, i.e. with the right to the protection of personal data. The reply to the last group of questions could have thus elaborated on this statement and clarified which personal data of the customers and recipients can be validly requested by credit institutions. However, in the case at hand BBVA denied that it requested the identification data of Safe's customers and of the recipients of the funds. It merely requested the identification data of Safe's agents who used BBVA's accounts. Moreover, the CJEU found the last group of questions not to be sufficiently precise because they only referred generally to the Data Protection Directive without specifying any of its provisions which could be relevant in this context. The part of the preliminary questions which related to the Data Protection Directive was therefore considered inadmissible.  


The replies of the CJEU to the preliminary questions point out in the direction of giving a certain degree of flexibility to the national legislators and to the institutions and persons which apply customer due diligence measures. On the other hand, the measures prescribed or authorised by the national authorities and the measures applied in individual cases by banks and other institutions and persons covered by the AML Directive need to be preceded by comprehensive risk assessments. Those risk assessments should lead to the definition of measures which are appropriate to the identified level of risk. The measures can vary depending on the type of customer, business relationship, product or transaction.

This kind of well-balanced approach seems in line with the objectives of the AML Directive and with the CJEU's case-law which recognised preventing and combatting money laundering and terrorist financing as an overriding reason in the public interest.

The CJEU added a further safeguard at the later stages of the judgment: the proportionality of the customer due diligence measures depends not only on the results of the risk assessment but also on their compliance with the fundamental rights and freedoms and general principles of law. The Court specifically mentions the principle of free competition and the right to the protection of personal data enshrined in Article 8 of the Charter.

In Safe the CJEU did not however provide any specific indications on the issue which personal data can be requested from the customer in the context of due diligence measures and in which circumstances. This was so because the last group of preliminary questions was based on facts which were disputed in the proceedings and eventually this last group was declared inadmissible by the Court.

The AML Directive does not really address the matter how the measures it designs relate to the protection of personal data. In fact, there is only one point in the text of the Directive which touches upon that issue. It is Recital 33 which refers to the applicability of national data protection laws and of the international transfers rules of the Data Protection Directive in the context of the transmission of information to the Financial Intelligence Units (FIUs) and the disclosure of information about such a transmission.

On the other hand, the new fourth Anti-Money Laundering Directive 2015/849 is much more outspoken in this respect. Its Chapter V implicitly states that Article 7(e) of the Data Protection Directive constitutes the legal basis for processing personal data for the purpose of the prevention of money laundering and terrorist financing by recognising, in Article 43, that such processing is a matter of public interest. The same Chapter deals also with the issue of the information that needs to be provided to the customer before establishing a business relationship or carrying out an occasional transaction. Finally, it lays down more precise indications with regard to the transmission of information to FIUs and to the disclosure of that fact to the customers. According to Article 41(4) this issue should be regulated in national laws which must strike the balance between the access of the customer to the personal data and the interests of the proper functioning of the anti-money laundering procedures and investigations.

The provisions on the different kinds of customer due diligence are also more precise in the new Directive. There is no longer a derogation from standard due diligence for financial institutions. The Directive is now accompanied by three annexes. The first of these annexes contains a non-exhaustive list of risk variables that shall be taken into account when determining the extent of customer due diligence measures. The second annex includes a non-exhaustive list of factors which point out to a potentially lower risk of money laundering and terrorist financing, i.e. the degree of risk that might trigger the application of simplified customer due diligence. Finally, the third annex is a non-exhaustive list of factors suggesting a higher degree of risk which requires the application of enhanced customer due diligence. Generally speaking, the factors included in the three annexes relate to types of customers, geographic areas, and particular products, services, transactions or delivery channels. In addition, Articles 17 and 18 of Directive 2015/849 envisage guidelines on the risk factors and the measures to be taken in situations of simplified customer due diligence and enhanced customer due diligence respectively. Such guidelines shall be issued by ESAs, i.e. the European Supervisory Authorities (EBA, EIOPA and ESMA) by 26 June 2017.

Photo credit:

[1] Ley 10/2010 de prevención del blanqueo de capitales y de la financiación del terrorismo.
[2] The Executive Service of the Commission for the Prevention of Money Laundering and Financial Crime of the Bank of Spain - Servicio Ejecutivo de la Comisión de Prevención de Blanqueo de Capitales e Infracciones Monetarias del Banco de España.

Testing EU experimentalist governance in the Telecoms sector

Marta Cantero (Postdoctoral Researcher, University of Helsinki | FiDiPro Project)

The building of a Digital Single Market for telecommunications is one of the main priorities of the current European Commission.[i] Yet, the achievement of an actual single market for telecoms is still far from becoming a reality. The designed (multi-level) regulatory model for telecommunications places sector-specific National Regulatory Authorities at the core in the system for the implementation of the EU regulatory framework for telecoms. Moreover, the enforcement of the framework corresponds to the Member States under the national procedural autonomy. However, in order to preserve the legal integrity of the EU rules, the consistent application of the EU telecoms framework and the achievement of its policy objectives builds on a sector-specific supervisory mechanisms that grants the European Commission greater powers to monitor the different regulatory approaches of national regulators: Articles 7 and 7a of the Framework Directive on telecoms regulation.[ii] These articles put in place a consultation and monitoring system of a post-legislative nature that aims at consolidating the internal market for telecoms based on a combination of hard and soft law techniques, but is this system up to the task?

This brief post examines a pending case (Case C28/15, Koninklijke KPN NV and Others v Autoriteit Consument en Markt) that highlights the deficiencies of the market-consolidating mechanism put in place.


In a nutshell, the case deals with a clash between the Dutch telecoms regulator, which issued a regulatory decision implementing a Commission Recommendation on termination rates,[iii] and the Dutch Trade and Industry Appeals Tribunal, which overruled the National Regulatory Authority’s (NRA) decision in the context of a procedure of judicial review following the appeal of the regulatory decision by some market telecoms operators. In particular, the national Court required the NRA to deviate from the Commission Recommendation on the grounds that there were no reasons for justifying a modification in the methodology used for calculating caps on termination rates.

However, this is the second time that the regulatory decision following the Recommendation is contested in front of the national court in The Netherlands. Already in 2010, the Dutch regulator (at that time OPTA, now ACM[iv]) issued a decision in line with the guidance provided in the Commission Recommendation on termination rates. That triggered an initial response by the telecoms operators, who appealed the regulatory decision. In very broad terms, and leaving aside further competition concerns and issues of market analysis that were also object of the plea, the Court, upholding the appeal, argued that despite the Commission’s Recommendation on termination rates, conditions on the national market remained unchanged and, therefore, there was no reason to adjust the methodology for cost calculation in accordance with the Recommendation. Essentially, the court concluded that the inefficiencies in retail pricing cannot be resolved by imposing a “more invasive measure” at wholesale level, given that the retail mobile market was already considered competitive.[v] As a result, the Court established new cap prices for termination rates and compelled the regulator to take a new decision setting the relevant rates on the basis of a different cost-methodology than that suggested by the European Commission. As part of the consultation procedure enshrined in Article 7a Framework Directive, the national regulator notified the European Commission the new decision compliant with the court’s judgment. In view of the Commission, that decision could create a barrier to the Internal Market. This led to the opening of a Phase II investigation under Article 7a procedure. Such a situation placed the NRA in the middle of a “tug-of-war” between the European Commission and the national judiciary. At that time, the national regulator could do anything but to give effect to the judgment of the highest administrative court in The Netherlands. However, two years later, in the context of a new market analysis, the regulator –perhaps pressured by the Commission’s investigation under the 7a procedure– issued a new decision following the European Recommendation. As expected, the new decision was again appealed in front of the national court. However, on this occasion, the national court decided to refer the case to the European court for preliminary ruling.

Issues at stake

In brief, the national judge asked the European court to clarify the discretion of the national judge[vi] to depart from a EU Recommendation on the basis national legal and factual(!) circumstances. The national court also seeks clarification as to the competence of the national court to assess the proportionality of the NRA’s performance within the context of the judicial review of regulatory decisions (Article 4 of the Framework Directive). Accordingly, the case addresses three fundamental legal (and institutional) tensions: 1) the legal and factual effect of soft-law; 2) the institutional and substantial limits of the judicial review of the activity of the national regulator; and 3) the proportionality of the NRA’s regulatory activity when giving effect to a EU Recommendation in a situation where the factual circumstances of a national market remain unchanged (reflecting a clash between the national regulator and the national judiciary).

So far, we do not have a final Judgment from Luxembourg. However, the analysis of the recently issued Opinion (28th April) already provides warnings about the institutional problems that this case entails, in particular, for those other NRAs in Europe that are facing a similar situation and that, therefore, are awaiting a decision.

AG Opinion

In the Opinion, AG Mengozzi holds that, despite its non-binding nature (para. 54), the national court has to “take into consideration” the Recommendation on termination rates (para. 57). Moreover, he also posits that the national judge must act with “extreme caution” when deciding to depart from the methodology suggested by the Commission (paras.53 and 64). Advocate General also considers that it is not a problem of incompatibility of the national law with the EU legal provisions. In particular, he states that he finds (para. 72)

“very difficult to conceive that the national law, as it has stated that court, namely, as proceeds from Union law, requires, by its wording and its capacity, departing from the calculation model recommended by the Commission”

However, and given that that does not mean that there is only one appropriate model to give effect to the provisions contained in the Access Directive[vii], Mengozzi acknowledges, the specific characteristics of the Dutch market could lead the national court to depart from the recommended model (para. 75).

The second part of the Opinion focuses on the assessment of proportionality of the regulatory decision in accordance to the regulatory objectives to be pursued by NRAs under the Framework Directive as part of the judicial review. In Mengozzi’s view, the scope of judicial control of the regulatory activity reaches the proportionality assessment (paras. 80 and 84). As to this proportionality assessment, he holds that, in his view, following the Recommendation would entail a presumption of proportionality with the EU regulatory objectives enshrined in Article 8 of the Framework Directive; namely, promotion of competition, contribution to the development of the Internal Market, and promotion of the interests of the citizens of the European Union. This is important because the regulator’s justification to impose an obligation in a regulated market (wholesale) was based on the effects to be produced on a non-regulated market (retail). Accordingly, when it comes to the burden of proof, and given that it would require the demonstration of an impossible (or excessively difficult to provide) evidence, he concluded that the national court cannot require the NRA to sufficiently prove the effective achievement of the regulatory objectives (paras. 92 and 96).


Harmonizing the Internal market under a multi-level governance structure is not an easy task, and this case overly illustrates the difficulties that such endeavor entails. The underlying issues that the case involves can be summarized as follows:

1.      First, should the European court follow AG’s Opinion it would mean that the national judge, when deciding on the appeal, can overturn the analysis performed by the regulatory authority, as it already did with the first national ruling; i.e. the national judiciary would be acting as a de facto regulator. This results in an institutional conflict that slows down the integration of the telecoms market –the case has been ongoing since 2010. In my view, this judicial spillover raises the question as to whether the intervention of the national judiciary into the regulatory activity needs to be balanced against the principles of equivalence and effectiveness in the context of the implementation of a non-binding instrument.
2.      Secondly, the case casts doubts on the effectiveness of the sector-specific supervisory mechanism put in place under Articles 7 and 7a procedures of the Framework Directive, and the limits and actual effect of soft law as an integration tool.
3.      Thirdly, the multi-level governance design upon which the sector is build raises the question as to whether the national court should be entitled to determine the effective influence of national regulatory measures beyond the domestic marketplace. If the ECJ agrees with the Opinion, it would be for the national judge to decide on the effect of a national measure on the Internal Market – something that should belong to the ECJ. This calls for further answers concerning the viability of the telecoms market, as a fast-paced market, to coexist with divergences in Europe or, rather, whether the regime should be upgraded (e.g. more formal powers to the European Commission or the creation of a European Telecoms Agency, something that has failed so far).
4.      Finally, the related problem of building a single market for telecoms under a multilevel governance system. Given the relevance of the case for other NRAs around Europe that are facing similar situations, most of the regulatory decisions from European NRAs involved investigations initiated by the Commission under the abovementioned Article 7a procedure are on hold until the case is decided.


The case addresses classic and timely questions about the role and legal effect of EU soft law. In particular, when it comes to the effectiveness of soft law mechanisms for market-integration purposes; which is perhaps the most interesting aspect of the case.

Whatever the final outcome will be, it will have remarkable consequences for the current configuration of the telecoms institutional and procedural framework. One possibility is that the European court does not allow the national judiciary to depart from the Recommendation. In such case, it would mean that there is no room for domestic adaptation and Article 7a procedure would then help to boost non-binding decisions from the Commission. However, the other possibility is that the court follows AG’s interpretation. In my view, allowing departure from the Recommendation would render Article 7a procedure ineffective, provided that the national court would define to what extent the effect on the Internal Market of a national regulatory decision is sufficient so as to justify a mandatory compliance with a non-binding European instrument.

Photo credit:

[i] Political Guidelines for the next European Commission – A New Start for Europe: My Agenda for Jobs, Growth, Fairness and Democratic Change (15 July 2014), Jean-Claude Juncker.
[ii] Directive 2002/21/EC of the European Parliament and of the Council of 7 March 2002 on a common regulatory framework for electronic communications networks and services (Framework Directive), as amended by Directive 2009/140/EC and Regulation 544/2009.
[iii] Without much elaboration on the technical details, termination rates are the rates which telecoms networks charge each other to deliver calls between their respective networks; i.e. how much mobile phone operators can charge to connect calls on each other’s networks. The Commission Recommendation aims at harmonizing the costing methodology used in the calculation of price caps for termination rates; Commission Recommendation (2009/396/EC) of 7 May 2009 on the Regulatory Treatment of Fixed and Mobile Termination Rates in the EU. OJ L 20.5.2009, pp. 67-74.
[iv] OPTA (Onafhankelijke Post en Telecommunicatie Autoriteit, "Independent Post and Telecommunications Authority", in English) was replaced by a single “super watchdog” body: the Netherlands Authority for Consumers and Markets (‘ACM’) after the merger of the Netherlands Competition Authority (NMa), the Netherlands Consumer Authority, and the Independent Post and Telecommunications Authority of the Netherlands (OPTA). ACM became operational as of 1st April 2013.
[v] CBb Judgment of 31st August 2011,
[vi] Interestingly, the national court poses question(s) of the legitimacy of the court to deviate from the Recommendation, but it does not refer to the NRA’s discretion to not follow the recommended costing methodology, which is the situation in some other Member States.
[vii] Directive 2002/19/EC of the European Parliament and of the Council of 7 March 2002 on access to, and interconnection of, electronic communications networks and associated facilities (Access Directive).

Thursday, 26 May 2016

The proposed new Audiovisual Media Services Directive: Key Features

Lorna Woods, Professor of Media Law, University of Essex

After a draft was leaked last week, the Commission proposal to revise the Audiovisual Media Services Directive (AVMSD) is now out.  Once again we see the Commission proposing the roll-out rather than the roll-back of regulation in the face of sector change.  The following provides an overview of some of the issues.

The first change is an extension of material scope.  The Commission explains in its Memo/16/1895 that a ‘limited extension’ will occur as the new proposal applies to ‘video-sharing platforms’, such as YouTube.  “Video-sharing platform services” are defined in new Article 1(aa) AVMSD (Art. 1(1)(b) of the proposal):

‘… a service, as defined by Articles 56 and 57 of the Treaty on the Functioning of the European Union, which meets the following requirements:
(i)                  the service consists of the storage of a large amount of programmes or user-generated videos, for which the video-sharing platform does not have editorial responsibility;
(ii)                the organisation of the stored content is determined by the provider of the service including by automatic means or algorithms, in particular by hosting, displaying, tagging and sequencing;
(iii)               the principal purpose of the service or a dissociable section thereof is devoted to providing programmes and user-generated videos to the general public in order to inform, entertain or educate;
(iv)              the service is made available by electronic communications networks within the meaning of point (a) of Article 2 of Directive 2002/21/EC.’

The phraseology and organisation here is different from the leaked draft. It makes clear the cumulative nature of the conditions but also clarifies that the organisational features of the video-sharing platforms identified are illustrative not an exclusive list.  It is also starting to engage with the issues surrounding editorial choice in an environment where ‘suggestions’ are made by programming – following big data profiling or just paid prominence. Moreover, the proposal integrates the point that such platforms can be caught if a ‘dissociable segment’ satisfies the definition, whereas the leaked version had a separate subclause (a ter) that applied a principal purpose test not just to video-sharing platforms but services defined in (1)(a)(i).

No doubt there will be much comment on the workability of this definition – not least where it draws the boundaries. Will there be difference in treatment between Instagram, Flickr and other photo-sharing sites, Twitter and Facebook (both of which have video capability, or link to videos) and Youtube, Vine and Vimeo; and are these sites similar to Dailymotion and For now, note the centrality of the concepts of ‘programme’, which by contrast to the leaked draft, gets a new definition (in Article 1(1)(b) AVMSD, replaced by Art. 1(1)(c) proposal), and ‘user-generated video’ (added to the AVMSD as (1)(ba)). This latter definition covers material created by end-users, but also material that such users may be re-using by uploading. This means the (unlawful) uploading of professional falls within the definition, but also material the creators of which are unknown.

The definition of programme does not apply just to video-sharing platform services, but is a central element in determining the scope of the AVMSD.  So, Rec. (3) (which was Rec 11 in the leaked draft) specifies that the AVMSD ‘should remain applicable only to those services the principal purpose of which is the provision of programmes to inform, entertain or educate’.  The purposes here are so broad that they can exclude nothing; the determinative element is therefore the programme.  This issue was the subject of litigation in the context of a press site which contained short video clips in New Media Online GmbH v. Bundeskommunikationssenat (Case C-347/14), in which the ECJ ruled that videos under a subdomain of a newspaper website could fall under the definition of a ‘programme’ within the AVMSD (an approach from which OFCOM has arguably differed in respect of its interpretation of the UK implementing regulations).   This position is reflected in Rec. 3, which notes that stand alone parts of newspaper sites can fall within AVMSD as can channels within video-sharing platforms.  Radio remains outside the AVMSD.

Under the current AVMSD,
(b)   ‘programme’    means    a    set    of    moving    images    with    or   without   sound   constituting   an   individual   item   within   a   schedule   or   a   catalogue   established   by   a   media   service   provider    and    the    form    and    content    of    which    are    comparable  to  the  form  and  content  of  television  broadcasting.   Examples   of   programmes   include   feature-length   films,   sports   events,   situation   comedies,   documentaries,   children’s  programmes  and  original  drama;

The proposal removes the phrase ‘and the form and content of which are comparable to the form and content of television broadcasting’ suggesting a move away from traditional television as the benchmark and towards a more open and arguably broader conception of just ‘an individual item’.

Rec 12 of the leaked draft is now found at Rec 26 and 27, slightly amended so as not to be limited to ‘video sharing’ platforms, though these are seen to raise particular issues.  Rec. 13 of leaked draft is now at Rec. 28 in a slightly amended form.  This change reflects the fact that the recitals refer to content restrictions rather than to scope, though Rec 28 contains the implicit acknowledgment that the proposal takes the possibility of regulation beyond those with editorial responsibility (even that at a very blunt level of choice – as in OTT services).  The Explanatory Memorandum skirts this issue, recognising that there will be a point of interplay with Articles 14 and 15 of the e-Commerce Directive (ECD). Those provisions provide immunity from damages for hosts with no knowledge of problematic content and prohibit the imposition of monitoring requirements (see also Rec. 30).  The proposal also recognises the need to include those services providers that are not established within the EU but are part of a group so as to ensure effectiveness of protection (Art. 28b).  In this context, we might be reminded of the reasoning of the Court in determining jurisdiction under the Data Protection Directive in Google Spain: legal form was not determinative of this question, but instead the business reality.  OFCOM in its response to the Commission’s consultation last year expressed concern about rules that would be ‘disproportionate and impractical’.  These provisions need also to be understood about the on-going trade negotiations with third countries, such as TTIP, which may affect their feasibility.

One of the main concerns with regard to video-sharing platforms is hate speech the understanding of which – in relation to all regulated platforms - ‘should, to an appropriate extent’ be aligned to Framework Decision 2008/913/JHA (concerning criminal expressions of racial hatred), specifically as regards the grounds on which hatred may be incited (Rec. 8).  What this means in practice, given the qualifiers used, as well as the relationship with the ECD in respect of video-sharing platforms, is uncertain (see Art. 28a(5)).  New Art. 6 AVMSD simply imposes on Member States the obligation to use ‘appropriate means’, the meaning of which is elaborated in Art. 28a. That provision points to a balancing of competing interests, which may allow for a certain degree of subjectivity and variation across Member States.  Art. 6a deals specifically with the protection of minors from a wider range of content – that likely to impair physical mental or moral development.   This envisages the need to give information to viewers so that they may make appropriate choices of viewing, rather than the imposition of technical measures. 

Another contentious issue has related to the country of origin (COO) principle, specifically where AVMS providers engage in forum shopping and ‘broadcast back’ to the ‘original’ target country.  This has always been problematic, with a body of jurisprudence on abuse of rights leading to specific exception provisions in the AVMSD.  The idea of COO, however, has always been popular with industry players as it avoids re-versioning costs and other costs associated with separate markets.  Whatever the view on COO, the anti-abuse provisions in AVMSD were complex and the issue of establishment open to interpretation.  The principle of freedom of re-transmission is restated but the possibility of derogating is extended to all audiovisual media services, not just broadcasting as is currently the case (see proposed Art. 3(2) – note differences in procedure between linear and non-linear services apply).  The grounds are those set out in Art 6, which contains an extended category of grounds of hate speech prohibited, and new Art. 12 which contains the ‘pornography provision’: transmission of relevant material is permitted, but in a way so that minors cannot access the material.  This applies to all AVMS providers.  The current broadcasting-only, protection of minors provisions (Art. 27 AVMSD), which currently act as triggers for the Art. 3 procedure, will be deleted.

While the AVMSD was a minimum harmonisation directive, recognised by Art. 4(1) AVMSD which allowed Member States to impose higher standards in respect of all fields coordinated by the directive, the proposal is now to limit the issues in respect of which Member States may impose stricter rules to Art. 5 (information obligations), 6 (hate speech), 6a (development of minors), 9 (standards for commercial communications), 10 (sponsorship), 11 (product placement), 12 (protection of minor- technical measures), 13 (on-demand quotas), 16 (tv European quotas), 17 (tv independent quotas), 19-26 (advertising and teleshopping rules), 30 (NRA) and 30a (ERGA).  In respect of the other provisions, it seems the AVMSD provides maximum standards.  It is notable that this latter category includes the provisions that are specific to video-sharing platforms as well as long-standing provisions such as the news reporting provisions.

There has been some ‘alignment’ of rules for linear and video on demand services (e.g. Art. 12).  This at an abstract level makes sense.  Commentators suggest that the industry trend is for entertainment, television and similar companies to focus on making and assembling content for distribution across the multiplicity of digital platforms available, in ways appropriate to those platforms but between which there may be overlap of form and content. Certainly, there is inter-platform competition.  So these changes are aimed at ensuring the mythical ‘level playing field’.  

The provision that has caught some attention when the proposal was leaked was that which imposes a European quota requirement on on demand AMS providers: at least 20% of the catalogue has to be European, and these works should be given prominence (Art. 13(1)).   It replaces the current provision which, in the words of the Commission ‘leaves room for testing different approaches’ but which potentially ‘unlevels’ the playing field.  Note that there is no ‘so far as practicable’ phraseology in this obligation (by contrast with the long-standing obligation on broadcasters), although member States may waive obligations in relation to small and micro enterprises (Art. 13(5)).  The obligation of ‘prominence’ is also not qualified (contrast the UK rules regarding ‘due prominence’ of PSB). Presumably it is intended at address the point made by the Society of Audiovisual Authors that currently on Netflix ‘where you have to look for European works (or even national works) under the rubric “Foreign Films”’ – not necessarily the most enticing branding.  

The current TV quotas rules have not addressed the problem of scheduling undermining the effectiveness of the quotas, a point noted in the response to the Commission’s consultation.   The definition of “European” has been left unchanged – as have the tv quotas.  This proposal will no doubt please the EU film industry, though it is likely to be less popular with the distribution sectors, which are already warning about strangling a still not mature industry.  Against this background it is noteworthy that Netflix has produced a series in Europe (Marseille – perhaps to get a stronger foothold in the French language market) and is about to launch a second, as well as engaging with local broadcasters (e.g. “Kiss Me First” with Channel 4; “Suburra” with RAI). 

The proposal also introduces a requirement for Member States to set up legally distinct and functionally independent regulators, in many aspects following the Recommendation of the Council of Europe (Rec (2000) 23). While the desirability of independent regulators is recognised in most Member States as a way of safeguarding freedom of expression while achieving other societal and political goals, there is no such obligation in the current AVMSD framework. The need to introduce such a requirement may be a response to developments in some of the Member States where there have been changes to the regulatory architecture in respect of the media with consequent concerns about media independence. It further specifies with a non-exhaustive list the remit on such regulators:  media pluralism, cultural diversity, consumer protection, internal market and the promotion of fair competition.  These roles must be established in law and carry with them enforcement powers.  A right of appeal for viewers/end-users must be provided.  Significantly, this requirement applies across all AVMS providers, including video-sharing platforms. 

The proposal also formalises the European Regulators Group for Audiovisual Media Services (ERGA) (which was established on the basis of a Commission Decision in 2014), in response to a perceived need for greater senior level cooperation in European audiovisual policy developments. The response to the group has been mixed, some questioning whether it adds anything to the existing groups, such as the Contact Committee and the European Platform of Regulatory Authorities (EPRA) which exists outside the EU framework. Alternatively, given the proposed expansion of the AVMSD and the uncharted territory awaiting the regulators, a mechanism for coordination may be important for the functioning of the COO principle.  ERGA has already produced reports for the Commission on independence; on the protection of minors in a converged environment; and on material jurisdiction as part of the preparation for the review of the AVMSD.

Historically, the broadcasting and now the audiovisual sector has revealed deep divides between member states and also between various sectors of industry.  The Commission has no doubt attempted to produce a balance of interests after an extensive review process.  What will remain once the Council and the European Parliament start to look at this, especially after what is likely to be intensive lobbying, is anybody's guess.  It may even be affected by Brexit; while the directive should be agreed before any UK exit, surely the UK’s negotiating position would be weakened between any ‘no’ vote and actual exit, shifting the balance between the free market and dirigiste Member States.

Barnard and Peers: chapter 14
Photo credit: Theon Greyjoy

Tuesday, 24 May 2016

EU law and the ECHR: the Bosphorus presumption is still alive and kicking - the case of Avotiņš v. Latvia

Stian Øby Johansen, PhD fellow at the University of Oslo Faculty of Law*

Yesterday, 23 May 2016, the Grand Chamber of the European Court of Human Rights (ECtHR) delivered its judgment in the case of Avotiņš v. Latvia. This seems to be the ECtHR’s first detailed appraisal of the so-called Bosphorus presumption (the rule on the relationship between EU law and the ECHR) after the Court of Justice of the European Union (CJEU) in Opinion 2/13 rejected a draft agreement providing for the accession of the EU to the European Convention of Human Rights (ECHR). It also provides a first glimpse of how the ECtHR views the EU law principle of mutual trust, which has become particularly dear to the CJEU over the last couple of years.


For the uninitiated: the Bosphorus presumption refers to a doctrine in the case-law of the ECtHR that goes back to the 2005 judgment in Bosphorus Hava Yolları Turizm ve Ticaret Anonim Şirketi v. Ireland. In that judgment the ECtHR first stated, in line with previous case-law, that member states of an international organization (such as the EU) are still liable under the ECHR for “all acts and omissions of its organs regardless of whether the act or omission in question was a consequence […] of the necessity to comply with international legal obligations” (Bosphorus para 153). It also recognized “the growing importance of international cooperation and of the consequent need to secure the proper functioning of international organisations” (Bosphorus para. 150). In an attempt to reconcile these two positions, the ECtHR established what is now known as the Bosphorus presumption or the presumption of equivalent protection of ECHR rights by the EU, even though the EU is not a party to the ECHR:

155. In the Court’s view, State action taken in compliance with such legal obligations is justified as long as the relevant organisation is considered to protect fundamental rights, as regards both the substantive guarantees offered and the mechanisms controlling their observance, in a manner which can be considered at least equivalent to that for which the Convention provides […]. By “equivalent” the Court means “comparable”; any requirement that the organisation’s protection be “identical” could run counter to the interest of international cooperation pursued […]. However, any such finding of equivalence could not be final and would be susceptible to review in the light of any relevant change in fundamental rights protection.
156. If such equivalent protection is considered to be provided by the organisation, the presumption will be that a State has not departed from the requirements of the Convention when it does no more than implement legal obligations flowing from its membership of the organisation.
However, any such presumption can be rebutted if, in the circumstances of a particular case, it is considered that the protection of Convention rights was manifestly deficient.

Many have been curious about whether the ECtHR would modify the Bosphorus presumption following the rather belligerent rejection of EU accession to the ECHR by the CJEU in Opinion 2/13. In the foreword of the ECtHR’s 2015 Annual Report its President, Guido Raimondi, indeed seemed to signal an interest in shaking things up (emphasis added):

The end of the year was also marked by the delivery on 18 December 2014 of the Court of Justice of the European Union’s (CJEU) eagerly awaited opinion on the draft agreement on the accession of the European Union to the European Convention on Human Rights. [T]he CJEU’s unfavourable opinion is a great disappointment. Let us not forget, however, that the principal victims will be those citizens whom this opinion (no. 2/13) deprives of the right to have acts of the European Union subjected to the same external scrutiny as regards respect for human rights as that which applies to each member State. More than ever, therefore, the onus will be on the Strasbourg Court to do what it can in cases before it to protect citizens from the negative effects of this situation.

Yet, in the ECtHR Grand Chamber judgment in the case of Avotiņš v. Latvia, it can clearly be seen that – spoiler alert – the Bosphorus presumption is still alive and kicking. Indeed, as I will show below, the ECtHR for the first time applies it to a case concerning obligations of mutual recognition under EU law. This is notable, since one of the main arguments the CJEU put forward in Opinion 2/13was that EU accession to the ECHR posed such a big threat to the principle of mutual trust that it would “upset the underlying balance of the EU and undermine the autonomy of EU law” (Opinion 2/13 para 194).


Before we look at how the Grand Chamber applied the Bosphorus, it is necessary to summarize the key facts of the case. Mr Pēteris Avotiņš is a Latvian national, who in May 1999 borrowed 100 000 US dollars from a company named F.H. Ltd. and undertook to repay that sum with interest by 30 June 1999. The loan contract was governed by Cypriot law, and Cypriot courts had non-exclusive jurisdiction to hear any disputes arising out of it.

In 2003, F.H. Ltd. brought proceedings against Avotiņš in a Cypriot district court, alleging that he had not repaid the above-mentioned debt. Since Avotiņš did not reside in Cyprus, notice of the proceedings and summons to appear had to be served on the applicants by Latvian authorities. There is some factual disagreement regarding the serving of this application (see para 19 of the judgment). It seems as if the summons slip had been signed, but the signature on the slip did not appear to correspond to the applicant’s name. Nevertheless, the Cypriot court ruled in Avotiņš’ absence on 24 May 2004, and ordered him to pay F.H. Ltd. 100 000 US dollar plus interest. According to the Cypriot judgment, the applicant had been duly informed of the hearing, but had not attended.

In February 2015, F.H. Ltd. applied to the Riga City District Court seeking recognition and enforcement of the Cypriot judgment. This request was first rejected, due to discrepancies regarding the postal address of Mr. Avotiņš. This rejection was appealed by F.H. Ltd. to the Riga Regional Court, which quashed the District Court’s rejection. Upon reexamination of F.H. Ltd.’s application by the District Court the application was granted in full – without the parties being present.

According to Avotiņš, it was not until 15 June 2016 that he became aware of the Cypriot judgment and the District Court order for its enforcement. He contacted the District Court immediately and acquainted himself with the Cypriot judgment and the Latvian order. Interestingly, before the ECtHR the Latvian authorities did not dispute these facts.

This is where things get complex (see paras 27-35 of the judgment), and I can for the sake of brevity only give a brief summary of the facts from this point out. First, Avotiņš did not attempt to appeal the Cypriot judgment. However, he decided to appeal the Latvian enforcement order on the grounds that it violated the Brussels I regulation (concerning jurisdiction over and recognition of civil judgments), which is part of EU law, as well as rules of Latvian civil procedure. Second, the Regional Court in October 2006 accepted Avotiņš’ submissions, and quashed the enforcement order. The District Court seemed to find that the Cypriot judgment was not enforceable due to the lack of the certificate referred to in Article 54 of the Brussels I regulation. Third, F.H. Ltd. appealed the October 2006 order of the Regional Court to the Supreme Court. At the start of the Supreme Court hearing in January 2007 F.H. Ltd. submitted copies of inter alia the certificate referred to in Article 54 of the Brussels I regulation. Later the same day the Latvian Supreme Court quashed the October 2006 order of the Regional Court, and ordered the recognition and enforcement of the Cypriot judgment. In doing so, the Supreme Court held that under article 36 of the Brussels I regulation a foreign judgment “may under no circumstances be reviewed as to its substance” (para 34 of the judgment, citing the January 2007 judgment of the Latvian supreme court).


The applications

Avotiņš then filed complaints against Latvia and Cyprus before the ECtHR. The application against Cyprus was rejected, due to being too late (see para 97 of the judgment, referring to a ECtHR decision of 3 March 2010). However, his application against Latvia was filed within the time-limits.

In his application against Latvia, Avotiņš argued that the Latvian Supreme court had infringed his right to a fair hearing, by recognizing and enforcing the Cypriot judgment which in his view was defective as it had been given in breach of his right to a defence. Several third parties intervened in the latter case, including the European Commission, which provided a lengthy submission on the applicability of the Bosphorus presumption to the case and the compatibility of Brussels I regulation with ECHR article 6 (the right to a fair trial).

The ECtHR’s introductory remarks

The judgment of the ECtHR, which was adopted by a majority of sixteen votes to one (with two judges appending a joint concurring opinion), opens with the premise that ECHR article 6 is applicable to the execution of foreign final judgments. According to the Court (para 98 of the judgment):

a decision to enforce a foreign judgment cannot be regarded as compatible with the requirements of Article 6 § 1 of the Convention if it was taken without the unsuccessful party having been afforded any opportunity of effectively asserting a complaint as to the unfairness of the proceedings leading to that judgment, either in the State of origin or in the State addressed.

The ECtHR then noted that it had “never previously been called upon to examine observance of the guarantees of a fair hearing in the context of mutual recognition based on European Union law” (para 98). However, the ECtHR added it had “always applied the general principle” that a request for recognition and enforcement of foreign judgments cannot be granted without the court examining the request “first conducting some measure of review of [the foreign] judgment in light of the guarantees of a fair hearing”.

Does the Bosphorus presumption apply?

Following these initial remarks, the ECtHR went on to consider whether and to what extent the Bosphorus presumption was applicable to the case. It did so over ten pages (paras 101-127), making this probably the longest treatment of this famed presumption by the ECtHR to day.

First, on the scope of the Bosphorus presumption, the ECtHR confirmed the principles laid down in its previous by referring to the summary of that case-law in paras 102-104 of its judgment in the Michaud case. From that case-law it follows that the substantive protection of human rights in the area of EU law that the Brussels I regulation belongs to is equivalent. In particular, this is confirmed by article 52(3) of the EU’s Charter of Fundamental Rights, which stays that the Charter has to be interpreted consistently with ECHR rights that correspond to it. The fundamental condition for applying the Bosphorus presumption was thus fulfilled.

Next, it follows from the ECtHR’s case-law that two further conditions must be satisfied for the Bosphorus presumption to apply. These are (1) the “absence of any margin of manouvre” on the part of the domestic authorities implementing an EU law obligation, and (2) the “deployment of the full potential of the supervisory mechanism” provided for under EU law. Applying these principles to the present case, the ECtHR first found that the Latvian Supreme Court did in fact not have any margin of manoeuvre in this case. In coming to this conclusion, the ECtHR pointed to the CJEU’s case-law on the relevant provisions of the Brussels I regulation, which “did not confer any discretion on the court from which the declaration of enforceability was sought” (para 106 i.f.).

The ECtHR’s discussion of the second condition, the deployment of the full potential of the supervisory mechanisms under EU law, was much more extensive. The Latvian Supreme Court had not requested a preliminary ruling from the CJEU regarding the interpretation of the relevant provisions of the Brussels I regulation. However, this was not decisive for the ECtHR, which stated (para 109):

this second condition should be applied without excessive formalism and taking into account the specific features of the supervisory mechanism in question. It considers that it would serve no useful purpose to make the implementation of the Bosphorus presumption subject to a requirement for the domestic court to request a ruling from the CJEU in all cases without exception […].

Following this statement, the ECtHR referred to cases where it has found that ECHR article 6 require domestic apex courts to give reasons when they refuse to refer questions to the CJEU for a preliminary ruling, “in light of the exceptions provided for by the case-law of the CJEU” (para 110). However, the ECtHR was quick to add that the review conducted in those cases differs from that in the present case, where “it examines the decision not to request a preliminary ruling as part of its overall assessment of the degree of protection of fundamental rights afforded by European Union law” (para 110).

For those reasons, the ECtHR found that “whether the fact that the domestic court hearing the case did not request a preliminary ruling […] is apt to preclude the application” of the Bosphorus presumption “should be assessed in light of the specific circumstances in each case” (para 111). It then pointed to the relevant circumstances at play in the present case: Avotiņš “did not advance any specific argument concerning the interpretation” of the relevant provisions of the Brussels I regulation, and he did not request that the Latvian Supreme Court should ask the CJEU for a preliminary ruling (para 111). Since there was thus norequest for a preliminary ruling, the fact that the Latvian Supreme Court did not ask for a preliminary ruling was not “a decisive factor” (para 111). Consequently, the ECtHR found that also the second condition for the application of theBosphorus presumption was satisfied.

Was the protection of ECHR rights “manifestly deficient”?

A finding that the Bosphorus presumption applies is not the end of it, however, since that presumption can be rebutted if the protection of the rights laid down in the ECHR was “manifestly deficient” in the present case (para 112). In the opening paragraph of this part of the judgment, the ECtHR points to the fact that the Brussels I regulation is based on the principle of mutual trust, and affirmed the importance of this principle in EU law (para 113):

The Court is mindful of the importance of the mutual recognition mechanisms for the construction of the area of freedom, security and justice referred to in Article 67 of the TFEU, and of the mutual trust which they require.

Nevertheless, the ECtHR soon went on to stress that the “methods used to create that area must not infringe the fundamental rights of the persons affected by the resulting mechanisms, as indeed confirmed by Article 67(1) of the TFEU” (para 114). This statement was immediately followed by some key critical remarks (para 114, emphasis added):

However, it is apparent that the aim of effectiveness pursued by some of the methods used results in the review of the observance of fundamental rights being tightly regulated or even limited. Hence, the CJEU stated recently in Opinion 2/13 that “when implementing EU law, the Member States may, under EU law, be required to presume that fundamental rights have been observed by the other Member States, so that …, save in exceptional cases, they may not check whether that other Member State has actually, in a specific case, observed the fundamental rights guaranteed by the EU” […]. Limiting to exceptional cases the power of the State in which recognition is sought to review the observance of fundamental rights by the State of origin of the judgment could, in practice, run counter to the requirement imposed by the Convention according to which the court in the State addressed must at least be empowered to conduct a review commensurate with the gravity of any serious allegation of a violation of fundamental rights in the State of origin, in order to ensure that the protection of those rights is not manifestly deficient.

By thus requiring domestic courts to presume the observance of fundamental rights by other member states, as the EU law principle of mutual trust requires, the domestic courts are “deprived of […] discretion in the matter, leading to automatic application of the Bosphorus presumption” (para 115). Although it is a bit difficult to discern exactly what the ECtHR is alluding to here, it is hard to disagree that the nature of the mutual trust principle creates a paradoxical situation (para 115 i.f.); a twofold limitation of the domestic court’s review of the observance of fundamental rights, due to the combined effect of the presumption on which mutual recognition is founded and the Bosphorus presumption of equivalent protection.

However, despite these apparent limitations on domestic courts when the principle of mutual trust is at play, the ECHR, which is a “constitutional instrument of European public order”, nevertheless requires of them to ensure that there is no manifest deficiencies (para 116, emphasis added):

"Accordingly, the Court must satisfy itself […] that the mutual recognition mechanisms do not leave any gap or particular situation which would render the protection of the human rights guaranteed by the Convention manifestly deficient. In doing so it takes into account, in a spirit of complementarity, the manner in which these mechanisms operate and in particular the aim of effectiveness which they pursue. Nevertheless, it must verify that the principle of mutual recognition is not applied automatically and mechanically […] to the detriment of fundamental rights – which, the CJEU has also stressed, must be observed in this context […]. In this spirit, where the courts of a State which is both a Contracting Party to the Convention and a Member State of the European Union are called upon to apply a mutual recognition mechanism established by EU law, they must give full effect to that mechanism where the protection of Convention rights cannot be considered manifestly deficient. However, if a serious and substantiated complaint is raised before them to the effect that the protection of a Convention right has been manifestly deficient and that this situation cannot be remedied by European Union law, they cannot refrain from examining that complaint on the sole ground that they are applying EU law."

The test laid down in the final sentence of the quoted paragraph is a tough one. It is therefore no surprise that Mr. Avotiņš was unable to meet its criteria.

What is more surprising, though, is how close he got to doing so. Although the ECtHR found the system of mutual recognition in the Brussels I regulation to be generally compatible with ECHR article 6 (paras 117-119), the ECtHR was skeptical about the Latvian Supreme Court’s interpretation and application of that regulation. Avotiņš had, as mentioned above, argued that the application for recognition of the Cypriot judgment should have been refused. According to the ECtHR he “raised cogent arguments in the Latvian courts alleging the existence of a procedural defect which, a priori, was contrary to [ECHR article 6] and precluded the enforcement of the Cypriot judgment in Latvia”. (para 120 i.f., emphasis added)

Moreover, the ECtHR found that the Latvian Supreme Court applied provisions of the Brussels I regulation that provided for exceptions to the obligation of mutual recognition too mechanically. The details here are quite technical, and concern the determination of the burden of proof – an issue that is not governed by EU law. In its concluding appraisal of the Latvian Supreme Court’s approach, the ECtHR stated (para 121): 'This approach, which reflects a literal and automatic application of Article 34(2) of the Brussels I Regulation, could in theory lead to a finding that the protection afforded was manifestly deficient such that the presumption of equivalent protection of the rights of the defence guaranteed by Article 6 § 1 is rebutted.'

This is as close to a finding of “manifest deficiency” as we have ever gotten in the ECtHR’s case-law – more on that later – but again the specific circumstances of the case came to the rescue. According to Cypriot law Avotiņš had a “perfectly realistic opportunity” of appealing the seemingly final judgment (para 122). That the applicant was unaware of this opportunity did not matter, as he when entering into a loan agreement should have “ensured that he was familiar with the manner in which possible proceedings would be conducted before Cypriot courts” (para 124).

Consequently, the judgment fizzles out with a finding that the protection of fundamental rights was not manifestly deficient, in the specific circumstance of the present case (para 125).


This judgment is notable for at least three reasons. First, it is notable for simple fact that it is the first time the Grand Chamber applies the Bosphorus presumption since Opinion 2/13. The judgment confirms that the presumption is still alive and well, as one could probably expect despite some murmuring from the ECtHR president.

Second, it is notable for being the first case where the ECtHR goes right up to the edge of finding that a “manifest deficiency” in the protection of fundamental rights has occurred, but then backing off at the last second because of a specific feature of the case at hand. As a side note, though, the ECtHR’s reasoning is less clear than one could have hoped for here. Since the burden of proof seems to be key to the outcome of the Latvian Supreme Court’s judgment, and this is an issue that is not regulated by EU law, one might have argued that the Latvian Supreme Court did in fact have some “margin of manoeuvre”. It seems as if it could have complied with both the obligation of mutual recognition and ECHR article 6 by modifying the Latvian rules on the burden of proof. The reason for the lack of clarity on the part of the ECtHR here may be caused by opaqueness of the Latvian Supreme Court’s reasoning; it “tacitly presumed either that the burden of proof laid with [Avotiņš] or that [a remedy against the Cypriot judgment] had in fact been available to the applicant” (para 121).

Third, the case is notable for being the first where Bosphorus presumption takes the principle of mutual trust head on. Particularly because that principle has been elevated to constitutional status by the CJEU over the last couple of years – with Opinion 2/13 as a major catalyst (see, particularly, Opinion 2/13 paras 191-194). The ECtHR’s judgment is wary of the dangers of mechanical application of mutual trust obligations, and reaffirms the principles laid down in Bosphorus. Despite some critical comments, my best guess is that the CJEU will see this judgment as something of an olive branch from the ECtHR. From the CJEU’s perspective the case is indeed welcome, as cases concerning the Dublin Regulation (e.g. M.S.S. v. Belgium and Greece), where the ECtHR have found that EU Member States violated ECHR article 3 by sending asylum seekers back to the first EU country they entered, were not been well received. However, one must not forget that there are important legal differences between cases such asM.S.S. and the present case of Avotiņš. Notably, the Dublin regulation does not – despite myths to the contrary – contain any obligation to send asylum seekers back to the first EU country they entered. In Avotiņš the situation is markedly different: there is seemingly a clear obligation on the Latvian authorities to recognize and enforce the Cypriot judgment. Although, admittedly, the ECtHR’s unclear reasoning concerning the Latvian rules on burden of proof makes this distinction a bit less clear.

Barnard & Peers: chapter 9
JHA4: chapter II:8
Photo credit:
*Reblogged with permission from the Øby-kanalen blog