Friday, 19 May 2017

Uber: a taxi service or an app? Analysis of an ECJ Advocate-General’s view



Lorna Woods, Professor of Internet Law, University of Essex

Case C-434/15 Asociación Profesional Elite Taxi v. Uber Systems Spain SL, Opinion of the Advocate General, 11 May 2017

This case is the first before the Court of Justice specifically on the sharing economy and the extent to which coordination via platform should be treated as removing unnecessary red-tape, or as seeking to avoid regulation in the public interest (in the form of concerns about passenger safety) as well as permitting unfair competition.  While the Commission seems in favour of the (unequal) sharing economy, Advocate General Szpunar sees the position a little differently.

FACTS

Spanish law envisaged that taxi firms, and transport intermediaries, should hold a licence.  Asociación Profesional Elite Taxi (APET) sought to challenge the use Uberpop, an app which allows non-professional private drivers to transport passengers using the drivers’ own cars, where neither Uber nor the drivers have the requisite licences.   Passenger users download the app from Uber and provide their bank details to Uber.  On receiving a request for a car, the app notifies drivers and calculates the fare (this latter based on distance but also demand for taxi services at that time).  The payment is made to Uber, which deducts a percentage and then pays the remainder to the driver.  To use the app as a driver, an individual must comply with Uber’s terms and conditions. APET sought a cease and desist order and a prohibition of future similar behaviour on the basis of unfair competition.  Uber resisted APET’s claims on the basis that it was not providing transport/taxi services but was rather a digital intermediary. 

QUESTION REFERRED

The national court referred the questions of how to classify Uber’s services to the Court of Justice. The answer would affect with EU derived legal regime would be applied to Uber, with the corollary that the State’s freedom to impose licensing requirements would correspondingly vary depending on which regime was held to be applicable.  In essence, the question was whether Uber fell within the provisions of the e-Commerce Directive (Directive 2000/31) as an information society service provider, or whether the Services Directive (Directive 2006/123) or the TFEU itself applied in this context.  Here, there is a distinction between a service general and a service in the field of transportation.

OPINION

The first phase of the opinion comprises some general remarks about the significance of the ruling and the impact of different types of competence on the outcome.  The Advocate General also assumed that the respondent in the case should be the Dutch company (Uber BV), which operates the app in the EU, rather than the Spanish company, Uber Spain, which is responsible for advertising.

The Advocate-General then moved on to consider the scope of the e-Commerce Directive, specifically the meaning of ‘information society services’ as defined in Article 2(a) of that Directive by reference to Article 1(2) of Directive 98/34 (the Directive on notifying new technical barriers to trade). Under Article 1(2), an information society must be a service provided for remuneration, at a distance, by electronic means and at the individual request of a recipient. In the view of the Advocate-General, the questions of whether there is a service provided for remuneration and at individual request appeared unproblematic [para 27], but questions arose as to the test of whether the service is provided at a distance by electronic means.  In the eyes of the Advocate General the problem related to the fact that what was in issue was a ‘composite service’ [para 28].

The Advocate-General emphasised that the definition concerned services “’entirely transmitted, conveyed and received by wire, by radio, by optical means or by other electromagnetic means’” [para 29, quoting 2nd indent of 2nd subparagraph of Article 1(2), Directive 98/34, emphasis in Opinion].  So, services not delivered by electronic means did not fall within the scope of the e-Commerce Directive; services which were incidental to such services would likewise not be liberalised by the e-Commerce Directive. Assuming that they did would undermine the perceived effectiveness of EU law [para 31]. Thus:

… an interpretation of the notion of information society services which brings online activities with no self-standing economic value within its scope would be ineffective in terms of the attainment of the objective pursued by Directive 2000/31. [para 32]

The Advocate General suggested that a composite service would be treated as an information society service in two circumstances:

-          where the two elements could be seen as economically independent of one another they would be treated separately for regulatory purposes – the electronic element likely falling with the eCommerce Directive; and
-          where the service  provided was substantially or predominantly provided by electronic means.

A common example of the first case would be a three party situation where an intermediary service provider facilitates a transaction between a user and an independent service provider/seller.  While the intermediary provides added value, the trader here pursues an independent business.  In a two party situation – where the intermediary provider is also the provider of a service not provided by electronic means, the two elements cannot be seen as separable; rather, they ‘form an inseparable whole’ [para 35]. 

In that instance, it will be necessary to see if the composite service falls within the second category; that is, whether the bundle falls within the eCommerce Directive or outside it. For determining the answer to this second question, the key element is where the economic value lies. So where the main component is performed online that service should be classified as an information society service (assuming the other elements of the test are met); conversely, where it is not then the service does not fall within the eCommerce Directive. The Advocate General, citing Ker-Optika (Case C-108/09) suggests that this test would be satisfied in the case of online sales (via the seller’s own website).  Delivery of goods is ‘simply the performance of a contractual obligation’ [para 36]. 

Applying these tests here, the Advocate General noted that Uber provided more than a matching service of passengers to taxi drivers. It sets down the essential characteristics of the service to be provided (eg quality and age of vehicle; drivers to have licences and no criminal record), it informs drivers where and when there are likely to be a high volume of trips and/or preferential fares, and it sets the prices.  Uber maintains indirect control over drivers through its ratings function.  Thus, in the view of the Advocate General,

‘… Uber exerts control over all the relevant aspects of an urban transport service ….  (…) Uber therefore controls the economically significant aspects of the transport service offered through its platform’ [para 51].

 While the Advocate General sought to distinguish this case from the cases concerning whether drivers are employees of Uber, on the basis of this indirect control the Advocate General concluded that:

 ‘Uber’s activity comprises a single supply of transport in a vehicle located and booked by means of the smartphone application and this service is provided, from an economic standpoint, [citations omitted] by Uber or on its behalf’ [para 53]. 

The Advocate General sought to distinguish the activity of Uber from intermediary services on the basis that Uber drivers do not carry out an independent activity. Instead, their activity exists solely because of the existence of the platform.  By contrast, flight or hotel booking systems are separate from the independent services operated by the hotels and airlines and for whom the websites are just one mechanism of advertising their services. Furthermore, it is the hotels and airlines which control the prices and the conditions on which their services are offered. Finally, a choice is offered to the user between hotels/airlines. 

In opposition to the Commission’s views on the sharing economy, the Advocate General did not think that the fact that Uber did not own the cars was determinative. Uber is more than a ‘mere taxi booking application’ [para 64].  Because of the extent of the innovation on the transport sector caused by the way the apps links drivers and passenger and the conditions on which it does this, ‘it is undoubtedly the supply of transport which is the main supply and which gives the services the economic meaning’ [para 64].  The supply of connection services is ancillary to this.

Having determined that Uber’s services do not fall within the eCommerce Directive, the Advocate General considered the Services Directive: Article 2(2)(d) specifies that the Services Directive does not apply to transport services.  The Advocate General confirmed that Uber’s taxi services were transport services in the context of Article 2(2)(d) Services Directive as recital 21 refers to ‘urban transport [and] taxis’ [cited para 68].  The service likewise falls within the exception to the Treaty rules on free movement of services (Article 58(1) TFEU) and therefore subject to the specific transport sector rules in Article 90 TFEU et seq.

The Advocate-General concluded by considering the position should Uber’s app be deemed to fall within the eCommerce Directive.  He noted that Member States would then be limited in terms of the conditions that they could apply to such a service; drivers however would still be subject to any relevant national regulation. The Advocate General argued that Uber nonetheless could be penalised for unfair competition as ‘it is responsible not only for the supply whereby passengers and drivers are connected with one another, but also for the activity of those drivers’ [para 86], whether or not the booking app were seen to be separate from the transportation service or not.  The Advocate General therefore proposed that the eCommerce Directive

‘does not preclude requirements relating to the activity of transport in the strict sense being established in national law or the imposition of penalties on Uber for failing to comply with those requirements, including by means of an injunction ordering it to discontinue the service’ [para 88].

Comment

This opinion will be grabbing the headline news because of the headline fact that – shock, horror- Uber is a taxi company.  The reasoning used is worth a little more attention, because our understanding of that reasoning, if the Court of Justice follows the same lines, will affect any wider ramifications for the ‘sharing’ economy more generally.  So while the Advocate-General starts his opinion by suggesting that the subject matter of the case is ‘narrow’ (para 2), the repercussions are potentially a little broader.  A recognition of this fact can be inferred by the approach of the Advocate General to the question of scope of the eCommerce Directive and the insistence that the eCommerce Directive regulates services that are entirely delivered by electronic means – the emphasis is that of the Advocate General (para 29), with the result that it cannot be said

That any trade-related online activity, be it merely incidental, secondary or preparatory in nature, which is not economically independent is, per se, an information society service (para 37).

The opinion re-iterates this point when considering the app as an information society service (see below). The significance of this is that the eCommerce Directive cannot be used to avoid regulation of the main service, at least to the extent that such regulation is not a barrier to trade unacceptable to EU law more generally and as exemplified by the Services Directive, just because some aspect of the business is on-line.  This leads to a second general point: the Opinion is noteworthy for the way it manoeuvred the circumstances of the case around to the twin obstacles to national regulation of the eCommerce Directive and the Services Directive, especially given that the Commission, in its Communication on the Collaborative Economy (COM(2016)356) seem to view the Services Directive in particular providing a basis for such services. 

The tests identified by the Advocate-General will not seem unusual – talking about whether the different elements of a composite service are severable or preponderantly one thing or another can be seen elsewhere, for example in the case of goods and services (an example of which is given by Advocate General Szpunar in his reference to Ker-Optika, para 36), or even questions of competence. 

A couple of points can be made here.  The first is that in some of the factors to be taken into account in determining independence, the Advocate General comes close to eliding the question of what is the nature of the service with the question of who is providing it. Many of the factors the Advocate General considered reflect the Commission’s Communication from last year. Of potentially more consequence is the impact considering as a factor the question of whether the service would exist without the app.  This can be seen by contrasting the position of Airbnb with Uber.  If we look at control, Uber is caught because of its control over key aspects such as access to passengers, and price (detailed in paras 43-51); on these considerations Airbnb which does not set price might not be caught. However, if the test is that ‘the activity exists solely because of the platform’ (para 56), for Airbnb as for Uber, the answer might well be ‘yes’.  Of course, Uber and Airbnb are different in that transport does not fall within the Services Directive but rather is dealt with by specific provisions within the TFEU.  Nonetheless, the recitals to the Services Directive specify that:

it does not apply to requirements, such as road traffic rules, rules concerning the development or use of land, town and country planning, building standards (Rec 9),

some of which may affect the running of hotels or B&Bs. Further, the directive does not apply to taxation (Article 2(3)), so for example, taxes on short term lets may be unaffected by the Services Directive (though dealt with under the TFEU).

A further point that is worthy of note is the consequence of seeing Uber and the drivers as providing separate services.  The requirement to have a licence to provide a connection service in the context of transportation would fall within the scope of the eCommerce Directive and would be caught be the prohibition on authorisations set out in Article 2(h)(i); the Advocate General was of the opinion that is was unlikely that any restriction on the electronic aspect of the service could be justified by considerations of public interest.  The prohibition does not, however, extend to the regulation of transport services - they are not provided by electronic means.  So the regulation of taxi services remains possible and, as noted, Uber remains responsible for the drivers’ activities (para 86).  Arguing Uber’s activity as whole should benefit from the liberalising principles in the eCommerce Directive would run the risk of undercutting any form of regulation ‘because all traders are currently in a position to offer services by electronic means….’ (para 87).  Could a similar argument be put forward in other sectors of the sharing economy?  In principle, yes, but presumably only where the platform has exerted Uber-like control over the actual provision of the services.

Even when the Grand Chamber Court of Justice has handed down its ruling, this will not be the end of the Uber saga.  Currently pending before the Court is a reference from France concerning the imposition of penalties on Uber for running an unlicensed taxi service: Case C-320/16 Criminal Proceedings against Uber France, which raises the question of whether France should have notified its rules as a technical regulation under Directive 98/34.

Barnard & Peers: chapter 14

Photo credit: boing boing

Thursday, 18 May 2017

New Developments in the context of the European Citizens’ Initiative: General Court rules on ‘Stop TTIP’



Anastasia KaratziaAssistant Professor in EU Law, Erasmus University Rotterdam and currently Visiting Research Fellow at the School of Law and Social Justice, University of Liverpool

Introduction

A few months ago, we saw the first annulment by the EU’s General Court of a Commission Decision refusing registration of a proposed European Citizens’ Initiative (ECI), in the case of Minority SafePack. Last week, there was an even bigger development in the case law of the General Court regarding the interpretation of the ECI’s legal admissibility test: in the Stop TTIP case[1] the Court annulled another Commission’s Decision, this time not on a procedural ground such as the one in Minority SafePack, but on the substantive ground that the Commission breached Article 11(4) TEU (which sets out the power to adopt the ECI law), and Articles 2(1) and 4(2)(b) of the ECI Regulation, which sets out one of the criteria for the legal admissibility test.

In Stop TTIP, the General Court clarified a matter of contention between ECI organisers / stakeholders and the Commission viz. the scope of an ECI and, more specifically, the way in which the Commission had limited the acceptable subject-matters for the purposes of registering an ECI. These limitations were stipulated in the Commission’s letter of response regarding the refusal of registration for the proposed ‘Stop TTIP’ Initiative, which was submitted for registration in July 2014. The Initiative proposed to cease the negotiations for the Transatlantic Trade and Investment Partnership agreement (TTIP) between the EU and US, and to prevent the conclusion of the Comprehensive Economic and Trade Agreement (CETA) between the EU and Canada.

In more detail, ‘Stop TTIP’ had invited the Commission to ask the Council to repeal its decision to authorise the opening of the TTIP negotiations under Article 218(2) TFEU (which is the legal rule on the process of the EU negotiating treaties). It also asked the Commission to submit a proposal for a Council decision not to conclude CETA. In September 2014, the Commission replied to the organisers that both their proposals had been rejected on the basis of Article 4(2)(b) in conjunction with Article 2(1) of the ECI Regulation, because they fell outside the framework of the Commission’s powers to submit a proposal for a legal act of the Union for the purpose of implementing the Treaties.

The Commission’s reply revealed two limitations on the scope of the ECI.[2] First, the Commission stipulated that an ECI cannot invite the Commission to adopt preparatory acts. The Commission argued in its reply that its proposals to the Council under Article 218 TFEU to authorise the opening of negotiations for international agreements were not proposals for legal acts. The Council Decisions authorising the opening of negotiations for an international agreement are preparatory acts that produce legal effects only between the EU and its Member States and between the EU institutions. Therefore, the Commission’s relevant proposals to the Council lacked legal effect against third parties. Accordingly, the position of the Commission was that ‘Stop TTIP’ was not proposing any legal acts for the purpose of implementing the Treaties and could not be registered. Second, the Commission declared that an ECI cannot invite the Commission to propose a decision not to adopt a legal act such as a proposal not to conclude CETA, or to refrain from proposing a legal act. Such a proposal ‘would not deploy any autonomous legal effect beyond the fact of the legal act at issue not being adopted.’ The negative nature of the ‘Stop TTIP’ proposals, together with the fact that it arguably did not propose ‘legal acts’ as required by Article 11(4) and Article 2(1) ECI Regulation, led to the refusal by the Commission to register it. Notably, the limitations imposed by the Commission are not clearly indicated in the ECI Regulation. Instead, they resulted from the Commission’s own interpretation of the ECI’s legal framework.

After the Commission’s rejection, the ‘Stop TTIP’ organisers followed a twofold course of action: they brought a case before the EU General Court, which is the first instance part of the Court of Justice of the European Union (CJEU), contesting the Commission’s decision to refuse registration of their Initiative, and they started what they named ‘a self-organised ECI’, which was a campaign to collect signatures outside the contours of the ECI’s legal framework. The campaign went on to collect more than 3 million signatures, which the organisers handed over to the Commission in October 2015.

It becomes apparent from the above overview that the significance of the General Court’s judgment in the Stop TTIP case does not derive only from the question of whether the specific Initiative was wrongly refused registration, but also from the question of whether the ECI’s scope to propose EU action was rightly limited by the Commission beyond what is explicitly written in the ECI Regulation. In this sense, the General Court’s judgment is a milestone both for the ECI organisers themselves and for the functioning of the ECI as a mechanism for citizens’ participation. This short commentary will touch upon the key aspects of the judgment.

The arguments of the parties

The applicants in the case made two main arguments. They claimed that the Commission (i) breached Article 11(4) TEU and Article 4(2)(b) of the ECI Regulation, and (ii) breached the principle of equal treatment (Article 20 of the EU Charter of Fundamental Rights) because it had registered in the past the ‘Swissout’ Initiative which had very similar objectives with ‘Stop TTIP’. The judgment focused on the first ground of review and did not deal at all with the second.

In support of their claim, the applicants brought forward three main arguments. Firstly, they argued that the Council’s Decisions authorising the conclusion of an international agreement under Article Article 218(5) TFEU is not a preparatory act. With regard to the Initiative proposals concerning the CETA negotiations, which were already taking place at the time of the request for registration, a Decision by the Council to the Commission not to conclude CETA would not be a preparatory act but an act with legally binding effects. Regarding the Initiative proposals concerning a proposal by the Commission to the Council to repeal the Decision authorising the negotiations for TTIP, such a Decision would result to the termination of the negotiations, and would have been final and legally binding. In any case, the scope of an ECI should not be limited to proposing legal acts with definitive, legally binding effects vis-à-vis third parties. Neither the background to the ECI Regulation, nor the ECI’s overall regulatory framework call for such a restrictive reading of the term ‘legal acts’ (para 12).

Secondly, the applicants argued against the Commission’s position that an ECI cannot concern acts that deploy legal effects only between the institutions concerned. For the purposes of the ECI, the term ‘legal act’ should be defined broadly in light of Articles 288 – 292 TFEU, and should include Commission’s Decisions that are outside the ordinary legislative process (para 13).

Thirdly, the applicants referred to the potentially ‘destructive effect’ of the proposed Initiative on the negotiations for TTIP and CETA. This alleged ‘destructive effect’ cannot be put forward as a ground for refusal under the rationale that the Initiative’s proposals did not have the purpose of implementing the Treaties. In the view of the applicants, ‘the right of citizens to participate in the democratic life of the Union includes the possibility of citizens acting with the purpose of modifying, reforming, ratifying, or asking for a partial or total annulment of EU law’ (para 14).

The Commission’s main counter-arguments supported the position expressed in its 2014 letter of response to the organisers. The Commission reiterated its position that the Council Decision to approve the opening of negotiations for an international agreement is only preparatory because it only produces legal effects between the two EU institutions. Based on a ‘systematic and teleological interpretation’ of Articles 2(1) and 4(2)(b) of the ECI Regulation, it can be concluded that an act of preparatory character falls outside the definition of a ‘legal act’ for the purposes of registering an ECI (para 19). This argument was further supported by the assertion that the notion of democratic participation in the EU refers to the participation of citizens only in matters which (potentially) fall under their legal sphere. Instead, the Council and the Commission enjoy sufficient democratic legitimacy to be the ones to adopt acts that affect the relationship between the EU institutions (para 20).

In addition, the Commission repeated its argument that an ECI cannot ask it not to propose a particular legal act or to propose a decision for the non-adoption of a legal act. Interestingly, it referred to Article 10(1)(c) of the ECI Regulation which deals with the final stage of the ECI process, whereby the Commission is obliged to issue a Communication setting out ‘the action it intends to take, if any’. From this, the Commission concluded that only ECIs that aim to the adoption of a legal act or to the repeal of an existing legal act can be registered. Otherwise, a declaration by the Commission that, as a response to an ECI, it does not aim to propose the adoption of a legal act would have excessively limited the Commission’s monopoly of legislative initiative. According to this argument, an ECI asking for the Council to repeal a Decision opening the negotiations or asking it not to conclude an agreement, would have been an ‘unacceptable interference’ in an on-going legislative procedure (para 21).

The judgment of the General Court

The General Court began with a reference to the ECI’s legal framework. It mentioned Article 11(4) TEU, and the ECI Regulation, specifically Article 2(1) (definition of the ECI), Article 4(2)(b) (the legal admissibility test), and Article 10(1)(c) (the obligation of the Commission to respond to a successfully submitted ECI) (paras 23-27). It then explained that the ECI organisers had not asked the Commission not to submit a proposal to the Council for the signing and conclusion of TTIP and CETA. Instead, the organisers asked the Commission to submit to the Council two proposals: (a) a proposal to recall the authorisation for the opening of negotiations for TTIP; and (b) a proposal not to authorise the signing of TTIP and CETA and thus not to conclude these agreements (para 28). As such, the Court also clarified that the current case did not contest the competence of the Commission to negotiate TTIP and CETA. Instead, it was a challenge to the reasons given by the Commission for the refusal of the proposal (para 29).

Subsequently, the Court specified that the Commission has the competence to act in the way asked by the applicants, i.e. to submit to the Council the two proposals (paras 30-32), and went on to deal with the question of whether these actions can be excluded from an ECI either because they are preparatory acts, or because they are not necessary for the implementation of the Treaty, as the Commission had argued (para 33).

On the definition of a ‘legal act’ for the purposes of an ECI, the Court sided with the applicants: the notion of ‘legal act’ in Article 11(4) TEU, and Articles 2(1) and 4(2)(b) of the ECI Regulation cannot be interpreted to include only final EU acts with legally binding effects vis-à-vis third parties. The Commission’s position is not justified by the letter of the law or by the overall purpose of these provisions. This was all the more so since the actions in question, which concerned the conclusion of an international agreement, fit squarely into the definition of a ‘Decision’ in accordance with Article 288(4) TFEU, as clarified in Case 114/12 Commission v Council. Besides, a broad interpretation of ‘legal act’ is mandated by the democratic principle on which the EU is founded (Article 2 TEU) (paras 35-37).

In addition, the Court rejected the Commission’s argument that the Initiative could not have been registered because the suggested actions did not aim to the implementation of the Treaties and thus were destructive to the law-making process. According to the Court, there is nothing in Article 11(4) TEU or Article 2(1) ECI Regulation indicating that citizens cannot act through an ECI in order to prevent the adoption of a legal act. Furthermore, the conclusion of TTIP and CETA would have modified the EU legal order. As such, by advocating to stop the two agreements, the ‘Stop TTIP’ organisers were actually acting for the implementation of the current Treaties (para 41). In any case, Initiatives that propose the non-signature and non-conclusion of an international agreement produce legal effects since they may prohibit the modification of EU law intended by the said agreement (para 43).

Lastly, even though the Court did not explicitly address the applicants’ second claim on the unequal treatment of their Initiative in comparison with the Swissout Initiative, it did address the paradoxical situation that resulted from the treatment of the two Initiatives. This paradox resulted from the fact that, according to the Commission’s interpretation, an ECI could propose the termination of an existing international agreement but not the termination of the negotiations towards such agreement. The Court took a citizen-friendly approach in saying that citizens should not be obliged to wait until an agreement is concluded before they can contest the conclusion of the agreement through an ECI (para 44). In this sense, the Court has put proposals asking for the termination of negotiations on a par with those asking for the opening of negotiations, and has interpreted the scope of the ECI as being capable of encompassing both type of proposals.

Commentary

I had commented on an earlier publication that the ‘Stop TTIP’ case was a good opportunity for the CJEU to step in and point out the correct interpretation of Article 4(2)(b) of the ECI Regulation regarding proposals concerning the conclusion of international agreements. It would seem that the General Court has seized that opportunity. The judgment widens the scope of the ECI by completely overruling the Commission’s interpretation of legal admissibility in the particular context. In this sense, the judgment is a positive and constructive development not only for the ECI organisers, who had been waiting for it for almost three years, but also for those interested in starting an ECI campaign on a topic related to an international agreement, as well as for ECI stakeholders who have been calling for a more flexible legal admissibility test.

What makes the case especially interesting is the extensive reliance of the Court on the nature of the ECI as a democratic participation mechanism that intends to foster democratic dialogue and give citizens the opportunity to address the Commission in order to request action. For instance, the Court implicitly rejected the Commission’s first argument that a potential breach of Article 11(4) TEU was irrelevant and that the only relevant legal text should be the ECI Regulation which is based on Article 24 TFEU and stipulates the details of the legal admissibility test. Both the Court’s interpretation of ‘legal acts’ for the purposes of registering an ECI (paras 35-36) and that of ‘implementing the Treaties’ (para 41) relies on a joint reading of Article 11(4) TEU and the relevant provisions of the ECI Regulation. The Court even considered the ECI in light of the fundamental principle of democracy as included in the Preamble of the Treaty and the EU Charter of Fundamental Rights in order to broaden the scope of the right to bring an ECI beyond the Commission’s delineation (para 37).

In addition, the Court has held a more restrictive view than the Commission on what is an ‘unacceptable interference with the adoption of a legal act’ when it comes to an ECI. According to the Court, the very notion of citizens’ participation in the democratic life of the EU - of which the ECI is part - includes the possibility to ask for the modification, as well as the partial or total repeal of legal acts. A true form of citizens’ participation in the democratic life of the EU should give the opportunity to citizens to obstruct, or interfere with, the adoption of a legal act. Since it is entirely up to the Commission to decide the follow-up of a successfully submitted ECI after the public hearing of that ECI (Article 10 ECI Regulation), it could not be said that the registration of ‘Stop TTIP’ would have been an unacceptable interference with the legislative process or that it would have breached the principle of institutional balance (paras 45-46). It would seem, therefore, that the Court has taken into consideration the overall discretion of the Commission at the end of the ECI process when interpreting the legal admissibility test, which takes place at the beginning.

Given that this is only the second time that the General Court annuls a Commission’s decision to reject a proposed ECI, the answer to the question ‘what happens now?’ is not entirely clear. After the Minority SafePack case, the Commission registered the part of the ECI that it considered admissible. As a response to the judgment, the Commission also issued a Decision elaborating on its reasons for only registering part of the ECI. The situation this time around is more complicated. As mentioned above, the ‘Stop TTIP’ organisers went ahead with collecting signatures despite the refusal of their ECI. Impressively, within one year (October 2014 – October 2015) the campaign collected around 3.3 million signatures, more than any of the formally registered ECIs. Subsequently, the organisers stated in their website: ‘we demand that the European Commission treat us like a regular ECI which means we expect an official response from the European Commission and a public hearing in the European Parliament.’ The Commission is now faced with interesting dilemmas: Will it register the ECI or pursue the case further by appealing before the European Court of Justice? If it does register the ECI, will it accept the collected signatures or will it oblige the organisers to start over? In its plans to propose revisions to the ECI in the near future, will the Commission try to overturn the new judgment – or accept and fully incorporate it?

The factor of time also makes the upcoming Commission’s response to this case particularly noteworthy. Between 2014 and 2017 we have seen major developments with regard to TTIP and CETA, including 15 negotiating rounds on TTIP up to October 2016 and a proposal in July 2016 by the Commission to the Council for the signature and conclusion of CETA. More recently, the European Parliament voted in favour of CETA after Wallonia nearly blocked the agreement. All of these developments are in fact the exact opposite of what the ‘Stop TTIP’ organisers were requesting in their proposal, which indicates the importance of momentum to an ECI’s overall success.

On a final note, I wonder what the implications of the General Court’s judgment are with regard to future ECIs relating to Brexit. It would seem that the judgment has opened the door to ECI proposals objecting to a possible future agreement on the UK-EU relationship, assuming that such an agreement will be eventually negotiated on the basis of Article 207 and 218 TFEU. Of course we have a long way to go before this issue even becomes relevant – if it ever becomes relevant at all. However, such a scenario would certainly open a new dimension to citizens’ participation and voice in the Brexit process. Meanwhile, let’s see how the Commission will respond to Stop TTIP and how the organisers will continue their campaign.

Photo credit: Stop TTIP
Barnard & Peers: chapter 24



[1] The judgment is not available in English yet. This commentary is based on my own translation from the Greek version and any translation errors are mine.
[2] I had elaborated on the Commission’s Decision in an older publication: A.Karatzia “The European Citizens’ Initiative in practice: Legal admissibility concerns” (2015) 40 EL Rev. 509, pp. 516-518

Wednesday, 10 May 2017

Think of the children: the ECJ clarifies the status of non-EU parents of EU citizen children living in their own Member State




Professor Steve Peers

What immigration rights do non-EU citizens have under EU law? There are three main areas of EU law that address this issue: EU immigration and asylum law; EU treaties with non-EU countries; and EU free movement law. The latter area of law is focussed on EU citizens’ right to move between Member States, and so only covers non-EU citizens if they are family members of EU citizens who have moved to another Member State. Those rules also apply by analogy where an EU citizen with a non-EU family member has moved to another Member State, then moved back to that citizen’s home Member State. (These are known as Surinder Singh cases: see this discussion of the ECJ’s most recent ruling on such cases, from 2014).

But what if an EU citizen has a non-EU family member, but has never moved to another Member State? Such cases fall outside the scope of EU free movement law. They will therefore in principle fall solely within the scope of national law, unless either EU immigration and asylum law or EU treaties with non-EU countries apply (they usually will not). But in a limited number of cases, there is a fourth category of EU law which might apply to them: EU citizenship law.

This principle was first set out in the 2011 judgment in Ruiz Zambrano, which concerned Belgian children living in Belgium with two non-EU parents. The ECJ ruled that expelling the non-EU parents would in effect would result in the departure of the children from the EU, thereby risking the ‘genuine enjoyment of the substance’ of those children’s EU citizenship rights.

Subsequent ECJ case law (discussed here) made clear that this principle is apparently restricted to the non-EU parents of EU citizen children living in their home State. Cases very similar to Zambrano – two non-EU parents of an EU child – are rare, because Member States now rarely, if ever, confer nationality upon children simply because they are born on the territory. However, there are rather more cases where: a) a home-State EU citizen marries a non-EU citizen, b) their child gets home-State citizenship because one of her parents is a home-State citizen; and c) the parents’ relationship ends.

In those cases, Ruiz Zambrano still potentially applies, as long as the non-EU parent is the ‘primary carer’ for the home-State EU citizen child. In that case, removing this parent to a non-EU country would in effect force the EU citizen child to leave the EU as well.  But when exactly does the ‘primary carer’ test apply? The ECJ clarified this issue in today’s important judgment in Chavez-Vilchez and others.

Judgment

Chavez-Vilchez and others concerned a number of non-EU parents of Dutch children in the Netherlands, who sought to argue that they were primary carers of those children, and so entitled to residence in accordance with the Ruiz Zambrano judgment. The Dutch government argued that they could not automatically be considered primary carers if it was possible for the other parent, ie the Dutch citizen, to take care of the children:

…the mere fact that a third-country national parent undertakes the day-to–day care of the child and is the person on whom that child is in fact dependent, legally, financially or emotionally, even in part, does not permit the automatic conclusion that a child who is a Union citizen would be compelled to leave the territory of the European Union if a right of residence were refused to that third-country national. The presence, in the territory of the Member State of which that child is a national or in the territory of the Union, as a whole, of the other parent, who is himself a Union citizen and is capable of caring for the child, is, according to the Netherlands Government, a significant factor in that assessment (para 66)

While the Court of Justice agreed that the non-EU parents could not automatically be considered as primary carers where the home state EU citizen child was dependent upon them, the Court’s approach was more open. It began by restating prior case law: the key issue was ‘who has custody of the child and whether that child is legally, financially or emotionally dependent on the third-country national parent’ (para 68). It then reiterated, following Zambrano, that dependency was particularly significant (para 69). Then it added new detail on how to assess dependency:

…it is important to determine, in each case at issue in the main proceedings, which parent is the primary carer of the child and whether there is in fact a relationship of dependency between the child and the third-country national parent. As part of that assessment, the competent authorities must take account of the right to respect for family life, as stated in Article 7 of the Charter of Fundamental Rights of the European Union, that article requiring to be read in conjunction with the obligation to take into consideration the best interests of the child, recognised in Article 24(2) of that charter (para 70).

For the purposes of such an assessment, the fact that the other parent, a Union citizen, is actually able and willing to assume sole responsibility for the primary day-to-day care of the child is a relevant factor, but it is not in itself a sufficient ground for a conclusion that there is not, between the third-country national parent and the child, such a relationship of dependency that the child would be compelled to leave the territory of the European Union if a right of residence were refused to that third-country national. In reaching such a conclusion, account must be taken, in the best interests of the child concerned, of all the specific circumstances, including the age of the child, the child’s physical and emotional development, the extent of his emotional ties both to the Union citizen parent and to the third-country national parent, and the risks which separation from the latter might entail for that child’s equilibrium. (para 71; emphases added)

The Court went on to answer questions from the national court about the burden of proof in Zambrano cases, which were connected with the substantive test to be applied. The Dutch government had argued:

…the burden of proof of the existence of a right of residence under Article 20 TFEU lies on the applicants in the main proceedings. It is for them to demonstrate that, because of objective impediments that prevent the Union citizen parent from actually caring for the child, the child is dependent on the third-country national parent to such an extent that the consequence of refusing to grant that third-country national a right of residence would be that the child would be obliged, in practice, to leave the territory of the European Union (para 74).

Although the ECJ accepted that the burden of proof lay upon the non-EU parent (para 75), it also ruled that national authorities ‘must ensure that the application of national legislation on the burden of proof’ in such cases ‘does not undermine the effectiveness’ of EU citizenship rights (para 76). This meant that the authorities had to make ‘the necessary inquiries’ to find out where the EU citizen parent lived, ‘whether that parent is, or is not, actually able and willing to assume sole responsibility for the primary day-to-day care of the child’ and whether the EU citizen child was dependent upon the non-EU parent (para 77).

In effect, the Court ruled that while the non-EU citizen must make a prima facie case, national authorities share some of the burden to investigate some aspects of the case. Again, the substantive test applicable is less stringent than urged by the Dutch government.

Comments

Today’s judgment clarified a number of issues relating to Zambrano cases, following on from last year’s judgments in CS and Rendón Marín (discussed here) which clarified when non-EU Zambrano parents could be expelled for public policy reasons. While the 2016 judgments referred to the child’s best interests, age, situation and dependency (referring to case law of the European Court of Human Rights), today’s judgment also refers to ‘physical and emotional development’, ‘emotional ties’ to both parents, and the effect of separation on the child. All of these are factors relating to the child, not to the non-EU parent; but all of them nevertheless concern the child’s links with that parent.  

The Dutch government’s desired focus on the capability of the EU citizen parent takes a back seat to the child’s best interests, as further elaborated by the Court. This will protect more non-EU parents, but in a differential way. Oddly, the Court’s case law does not take express account of situations of joint custody, or the more general argument that the child’s best interest will usually be to maintain strong relationships with both parents (assuming they are not negligent or abusive).

Could it also be argued that the requirement of always seeking to identify a ‘primary carer’ is problematic from the point of view of gender equality?  Due to the division of labour relating to child care in practice, the Court’s rulings would classify more non-EU mothers than non-EU fathers as ‘Zambrano carers’; but the expulsion of those fathers will only increase the childcare demands on the EU citizen mother who remains, as well as disrupt the child’s right to maintain a relationship with his father. Of course, the presence of the parent who looks after a child day-to-day is essential; but children love the parent who kicks the ball as well as the parent who cooks the meal.

The procedural aspects of the Court’s judgment are interesting, but raise further questions: is there a right to appeal, to a decision within a reasonable time, to a lawyer, to legal aid? In last year’s judgments, the Court of Justice referred to concepts from EU free movement law and its relevant case law when discussing the substantive test for expelling Zambrano carers; but it made no such cross-references today. The long-term immigration status of the parent is also an open question, although Zambrano noted that there should be access to employment to make the residence rights of the parent effective.

Finally, a Brexit point: the draft EU position for negotiating acquired rights does not appear to cover Zambrano carers. From a technical point of view, this is logical because the case law concerns (from the UK’s perspective) non-EU parents of UK citizens who have not moved within the EU. So no free movement rights have been acquired; we are rather talking of EU citizenship rights which will necessarily be lost when the UK ceases to be a Member State, since citizenship of the EU is defined as deriving from the nationality of a Member State. But from a human point of view, any deterioration in legal status could damage or even shatter the family lives of the children concerned. Zambrano carers should therefore be protected ideally in the Brexit talks, or failing that by the UK unilaterally.


See also further reading on UK Zambrano case law by Charlotte O'Brien and Desmond Rutledge

Barnard & Peers: chapter 13
JHA4: chapter I:6
Photo credit: German embassy London

Friday, 5 May 2017

Immigration detention and the rule of law: the ECJ’s first ruling on detaining asylum-seekers in the Dublin system




Tommaso Poli, LL.M. candidate in Human Rights and Humanitarian Law at the University of Essex, School of Law.

One of the most controversial issues in immigration law is the detention of asylum-seekers. This issue was not initially addressed by the European Common Asylum System (CEAS), but is now addressed in some of the second-phase CEAS measures (the CEAS consists of the Asylum Procedures Directive, the Reception Conditions Directive, the Qualification Directive, the Dublin Regulation and the EURODAC Regulation).

In particular, the second-phase CEAS measures contain detailed rules on detaining asylum-seekers in two cases:  a) general rules in the Reception Conditions Directive, which were the subject of a first ECJ ruling in 2016 (discussed here) and a recent opinion of an Advocate-General; and b) more specific rules in the Dublin III Regulation, applying to asylum-seekers whose application is considered to be the responsibility of another Member State under those rules. Recently, the ECJ ruled for the first time on the interpretation of the latter provisions, in its judgment in the Al Chodor case.

As we will see, the Court took a strong view of the need for the rule of law to apply in detention cases. Moreover, its ruling is potentially relevant not just to Dublin cases, but also detention of asylum-seekers and irregular migrants in other contexts too.

The rules on detaining asylum-seekers in the context of the Dublin process are set out in Article 28 of the Dublin III Regulation. First of all, Article 28(1) states that asylum seekers can’t be detained purely because they are subject to the Dublin process. Then Article 28(2) sets out the sole ground for detention: when there is a ‘significant risk of absconding’. If that is the case ‘Member States may detain the person concerned in order to secure transfer procedures in accordance with’ the Dublin rules, ‘on the basis of an individual assessment and only in so far as detention is proportional and other less coercive alternative measures cannot be applied effectively’.

Next, Article 28(3) sets out detailed rules on time limits for ‘Dublin detention’; these are the subject of the pending Khir Amayry case. Finally, Article 28(4) states that the general rules on guarantees relating to procedural rights and detention conditions set out in the Reception Conditions Directive apply to asylum-seekers detained under the Dublin rules.

Al Chodor concerned the interpretation of the grounds for detention under Article 28(2): what is a ‘serious risk of absconding’?  The Dublin III Regulation offers some limited clarity, defining ‘risk of absconding’ as ‘the existence of reasons in an individual case, which are based on objective criteria defined by law, to believe that an applicant or a third country national or a stateless person who is subject to a transfer procedure may abscond.’ (Article 2(n) of the Regulation). 

Facts

The case relates to an Iraqi man and his two minor children who were travelling from Hungary in the Czech Republic, without any documentation to establish their identity, with the aim of joining family members in Germany. After stopping the Al Chodors, the Czech Foreigners Police Section (FPS) consulted the Eurodac database and found that they had made an asylum application in Hungary. As a consequence, the Al Chodors were subjected to the transfer procedure according to Article 18(1)(b) of the Dublin III Regulation. In addition, the FPS took the view that there was a ‘serious risk of absconding’, given that the Al Chodors had neither a residence permit nor accommodation in the Czech Republic, while they were waiting for their transfer to Hungary.

So, they placed the Al Chodors in detention for 30 days pending their transfer pursuant to Paragraph 129(1) of the national law on the residence of foreign nationals, read in conjunction with Article 28(2) of the Dublin III Regulation. The Al Chodors brought an action against the decision ordering their detention to the regional Court, which annulled that decision, finding that Czech legislation does not lay down objective criteria for the assessment of the risk of absconding within the meaning of Article 2(n) of the Dublin III Regulation. That Court accordingly ruled that the decision was unlawful. Following the annulment of the decision of the FPS, the Al Chodors were released from custody.

The FPS brought an appeal on a point of law before the Supreme Administrative Court against the decision of the Regional Court. According to the FPS, the inapplicability of Article 28(2) of the Dublin III Regulation cannot be justified by the mere absence in Czech legislation of objective criteria defining the risk of absconding. That provision subjects the assessment of the risk of absconding to three conditions, namely an individual assessment taking account of the circumstances of the case, the proportionality of the detention, and the impossibility of employing a less coercive measure. The FPS has submitted that it satisfied those conditions.

The Supreme Administrative Court was uncertain whether the recognition by its settled case-law of objective criteria on the basis of which the detention of persons pursuant to Paragraph 129 of the Law on the residence of foreign nationals may be carried out can meet the requirement of a definition 'by law' within the meaning of Article 2(n) of the Dublin III Regulation, in so far as that case-law confirms a consistent administrative practice of the FPS which is characterised by the absence of arbitrary elements, and by predictability and an individual assessment in each case. So the Court decided to refer to the European Court of Justice for a preliminary ruling asking whether Article 2(n) and Article 28(2) of the Dublin III Regulation, read in conjunction, must be interpreted as requiring Member States to establish, in a national law, objective criteria underlying the reasons for believing that an applicant for international protection who is subject to a transfer procedure may abscond, and whether the absence of those criteria in a national law leads to the inapplicability of Article 28(2) of that regulation.

Judgment

The Court of Justice first of all ruled that Article 2(n) of the Dublin III Regulation explicitly requires that objective criteria defining the existence of a risk of absconding be defined by the national law of each Member State (paragraph 27-28). Then, determining whether the word ‘law’ must be understood as including settled case-law, the Court reaffirmed that in interpreting a provision of EU law, it is necessary to consider not only its wording but also the context in which it occurs and the objectives pursued by the rules of which it forms part (judgment of 26 May 2016, Envirotec Denmark, paragraph 27).

So with regard to the general scheme of the rules of which Article 2(n) of Dublin III Regulation forms part, the Court, referring to recital 9 of that regulation, states that the regulation is intended to make necessary improvements, in the light of experience, not only to the effectiveness of the Dublin system but also to the protection of fundamental rights afforded to applicants under that system. This high level of protection is also clear from Articles 28 and 2(n) of that regulation, read in conjunction. As regards the objective pursued by Article 2(n) of the Dublin III Regulation, read in conjunction with Article 28(2) thereof, the Court recalls that, by authorizing the detention of an applicant in order to secure transfer procedures pursuant to that regulation where there is a significant risk of absconding, those provisions provide for a limitation on the exercise of the fundamental right to liberty enshrined in Article 6 of the Charter.

In that regard, it is clear from Article 52(1) of the Charter that any limitation on the exercise of that right must be provided for by law and must respect the essence of that right and be subject to the principle of proportionality. Furthermore, it is worth noting that in this ruling the European Court of Justice explicitly aligns its interpretation to the European Court of Human Rights (ECtHR), reaffirming that any deprivation of liberty must be lawful not only in the sense that it must have a legal basis in national law, but also that lawfulness concerns the quality of the law and implies that a national law authorizing the deprivation of liberty must be sufficiently accessible, precise and foreseeable in its application in order to avoid risk of arbitrariness (judgment of the European Court of Human Rights of 21 October 2013, Del Río Prada v Spain, paragraph 125).

The Court then concluded by stating that taking account of the purpose of the provisions concerned, and in the light of the high level of protection which follows from their context, only a provision of general application could meet the requirement of clarity, predictability, accessibility and, in particular, protection against arbitrariness. It follows that Article 2(n) and Article 28(2) of the Dublin III Regulation, read in conjunction, must be interpreted as requiring that the objective criteria underlying the reasons for believing that an applicant may abscond must be established in a binding provision of general application. In the absence of such criteria, the detention was unlawful.

Comments

First of all, the Court’s ruling is likely relevant to the interpretation of other EU measures concerning immigration detention. In the Returns Directive, which inter alia concerns the detention of irregular migrants (as distinct from asylum seekers), the ‘risk of absconding’ forms part of the ground for detention (as well as one of the grounds for refusing to allow the irregular migrant a period for voluntary departure); and it is defined exactly the same way as in the Dublin III Regulation. As for asylum seekers who are detained on grounds other than the Dublin process, a ‘risk of absconding’ is an element of one of the grounds for detention under the Reception Conditions Directive, but is not further defined. But a recent Advocate-General’s opinion notes (at para 73) that this clause aims to prevent ‘arbitrary’ detention, which was a key feature of the reasoning in the Al Chodor judgment. This surely points to a consistent interpretation of the two asylum laws. It follows that arguably the Court’s judgment should be relevant not just to Dublin cases but to any immigration detention of non-EU citizens in any Member State bound by the relevant EU legislation.

Secondly, this ruling has reiterated the principle by which although regulations generally have immediate effect in national legal systems without it being necessary for the national authorities to adopt measures of application, some of those provisions may necessitate, for their implementation, the adoption of measures of applicability by the Member States (judgment of 14 April 2011, Vlaamse Dierenartsenvereniging and Janssens, paragraphs 47 and 48).

Most significantly, the Court has reaffirmed the primacy of Human Rights law in EU asylum law implementation, highlighting that the development of the EU asylum law itself depends on its compliance with Human Rights law. In particular, the ECJ’s ruling in this case first of all reflects the ECtHR’s interpretation of the ‘arbitrariness’ of detention, which extends beyond the lack of conformity with national law. Notably, it states that a deprivation of liberty that is lawful under domestic law can still be arbitrary and thus contrary to the general principles, stated explicitly or implied, in the Convention (judgment of the European Court of Human Rights of 9 July 2009, Mooren v. Germany, paragraphs 73-77).

The Court’s ruling also reflects UN human rights norms. The Human Rights Committee’s General Comment No. 31 related to the nature of the general legal obligation imposed on State parties to the UN Covenant on Civil and Political Rights, which all EU Member States are State parties to, which reads that ‘in no case may the restrictions be applied or invoked in a manner that would impair the essence of a Covenant right’ (paragraph 4). Furthermore, the Human Rights Committee’s General Comment No. 35 points out that “arbitrariness is not to be equated with ‘against the law’, but must be interpreted more broadly to include elements of inappropriateness, injustice, lack of predictability and due process of law, as well as elements of reasonableness, necessity and proportionality” (paragraph 12, see also HRC, Van Alphen v. Netherlands, paragraph 5.8).

Finally, the Court’s ruling has confirmed the constitutional value of the Charter of Fundamental Rights of the European Union, which assumes a critical value in this historical period, since, as with any constitutional instrument, the more society as a whole is going through difficult times (such as the perceived ‘migration crisis’ in Europe), the more important it is to reaffirm its principles and values.

Likewise Article 52 of the EU Charter states that in no case may restrictions be applied or invoked in a manner that would impair the essence of a Charter right; in the context of detention, a fortiori it can be also affirmed that essential elements of guarantee for that right, as the requirement of lawfulness and non-arbitrariness for the right of liberty, cannot be disregarded in any circumstance. The Al Chodor ruling puts meat on the bones of that fundamental principle.

Barnard & Peers: chapter 26
JHA4: chapter I:5

Photo: Amygdaleza detention centre in Greece, credit: www.metamute.org

Tuesday, 4 April 2017

Guide to the Brexit Negotiations




Professor Steve Peers

Last week the Brexit process formally got underway, as the UK formally sent the ‘Article 50’ withdrawal letter to the EU on March 29, and the EU in turn drew up a draft of its Brexit negotiating guidelines.

The following is a detailed annotation of the text of the EU’s draft guidelines, which I compare throughout to the UK position – which is most fully set out in Theresa May’s Chatham House speech in January (discussed here). The draft guidelines might be changed before final adoption (due for April 29), but at present it seems unlikely there will be radical changes (if any), so my analysis is based on the text as it now stands. I will update this blog post if there are significant amendments.

I also make reference to the draft position of the European Parliament (EP). This is less important than the draft EU position, since the European Parliament is not (even partly) the negotiator in the talks – no matter how much it might claim to be, or how much some in the UK media believe that it is.  However, the EP position is still worth considering because the EP has a veto over the final Article 50 deal, and this power could influence the actual negotiators’ position. (There’s a full discussion of the EP’s role here). My comments are both legal and political – although we should keep in mind that most legal analysis about Article 50 TEU (which sets out the process of leaving the EU) is purely speculative in the absence of any ECJ case law.

While the initial attention in the UK has focussed on a misunderstanding of what the draft EU guidelines say about Gibraltar, the most significant issue is actually that the EU and UK in principle have many negotiating objectives in common. Most notably, the EU has accepted the UK’s objective of aiming towards a comprehensive EU/UK free trade agreement (FTA) without provisions on the free movement of persons or contributions to the EU budget.   

The devil will therefore be in the considerable details. For example, the EU and UK still disagree on the timing of Brexit talks, (possibly) the role of the ECJ, financial issues (the ‘divorce bill’), the Gibraltar issue (although this will be an issue for bilateral talks with Spain), whether the UK should comply with environmental and other standards as part of a deal, and whether ‘sectoral’ deals are possible. Further points of detailed disagreement will surely emerge as the talks get underway.

Is an EU/UK relationship based on a free trade deal the right way forward? Frankly, in my view, it’s not the first or even the second best option. But it is still a vastly better option than reversion to trade with the EU on a ‘WTO-only’ basis, as some are contemplating. This would lead to significant tariff and non-tariff barriers on many goods and services traded, and therefore hurt both sides.

There’s a democratic dimension to this too. Many of those voting ‘Leave’ were particularly concerned about EU budget contributions and the free movement of people, and an UK/EU FTA would resolve both concerns. But equally the ‘Leave’ side argued repeatedly that any concerns about a drop in UK trade with the EU were misplaced, because the UK and the EU would sign a free trade agreement without commitments relating to the free movement of persons or EU budget contributions.

They also claimed that such an agreement could be reached quickly and easily, on the UK’s terms. This was an obvious falsehood, as the EU’s draft negotiating guidelines have confirmed.  But rather than focus on this untruth, I believe we should move forward. It’s up to the government to move on with the negotiations, never forgetting that a reversion to a WTO-only trading relationship with the UK’s largest trading partner would both make a nonsense of the government’s aim of a free-trading ‘Global Britain’ and lack the slightest shred of democratic legitimacy. Moreover, it would damage both sides, and in particular put many people in the UK out of a job. These negotiations won’t be easy, but they are necessary. The time for fantasies, flag-waving and faffing about is over.

Annex I

On 29 March 2017, the European Council has received the notification by the United Kingdom of its intention to withdraw from the European Union and Euratom. This allows for the opening of negotiations as foreseen by the Treaty.

Note that the guidelines refer throughout to ‘negotiations’. Some have erroneously asserted that the EU will draft a text and present to the UK on a ‘take-it-or-leave-it’ basis. This is incorrect: Article 50(2) TEU refers several times to ‘negotiations’, and the draft guidelines reflect this.

The guidelines do not refer to any possible withdrawal of the UK notification, although the EP resolution (point L) assumes that a withdrawal of the notification is possible with consent. There is also an argument that the time period for withdrawal in Article 50 could be extended indefinitely. I will return to this issue another time.

Some have speculated that the EP might insist that the UK hold a referendum result on the outcome of the talks, as a condition for its consent to the deal. This would almost certainly backfire spectacularly, and in any event the draft EP resolution does not address this.

European integration has brought peace and prosperity to Europe and allowed for an unprecedented level and scope of cooperation on matters of common interest in a rapidly changing world. Therefore, the Union's overall objective in these negotiations will be to preserve its interests, those of its Member States, its citizens and its businesses.

The United Kingdom's decision to leave the Union creates significant uncertainties that have the potential to cause disruption, in particular in the UK but also in other Member States. Citizens who have built their lives on the basis of rights flowing from the British membership of the EU face the prospect of losing those rights. Businesses and other stakeholders will lose the predictability and certainty that come with EU law. With this in mind, we must proceed according to a phased approach giving priority to an orderly withdrawal.

The EU expressly asserts here its goal of an ‘orderly withdrawal’. Indeed Article 50(2) creates an obligation to negotiate with the withdrawing Member State (implicitly in good faith, as with any other negotiation under international law). Legally the wording of Article 50 equally suggests it aims to ensure an orderly withdrawal, and the ECJ is very likely to interpret it the same way. This necessarily implies that the EU cannot simply walk away from negotiations. However, in the event of a legal challenge it is unlikely (based on its prior case law on the EU’s international relations, which emphasises political discretion for the EU institutions) that the ECJ would rule that the general obligation to negotiate requires the EU to adopt (or refrain from adopting) specific negotiation objectives.

Politically this paragraph (and other similar references) confirms that the EU is aiming to reach a deal with the UK. So it would be false for anyone to assert that it is not.

In these negotiations the Union will act as one. It will be constructive throughout and will strive to find an agreement. This is in the best interest of both sides. The Union will work hard to achieve that outcome, but it will prepare itself to be able to handle the situation also if the negotiations were to fail.

The first sentence confirms that the EU will negotiate as a bloc, as Article 50 provides for (‘the Union shall negotiate and conclude an agreement with that State’). This does not preclude informal bilateral talks between the UK and key Member States’ governments, which is common when the EU discusses trade deals with non-EU states.

The overall tone here is diplomatic and asserts an intention to reach a deal, although the final sentence also asserts that the EU will prepare itself for a ‘no deal’ scenario as an alternative. This would likely entail getting draft EU laws ready for approval and gearing up national customs administrations et al for a UK departure presumably in one of two scenarios: a) on the date set out pursuant to Article 50; b) if the UK purports to leave the EU in breach of the terms of Article 50. In the latter scenario the EU Member States might not wait for the end of the Article 50 period, but declare that the UK has committed a material breach of the EU Treaties within the meaning of Article 60 of the Vienna Convention on the Law of Treaties, and suspend the UK from the Treaties.

The guidelines make no mention of amending the Treaties to remove reference to the UK, but one would not expect them to, since logically that is not an issue which the EU would negotiate with the UK.  It is not yet clear whether the EU will embark upon such a course; legally it might be argued that Article 50 is ‘self-executing’ in the sense that references to the UK automatically become legally irrelevant on Brexit Day. One issue might be whether other changes to the Treaties might be made at the same time; if so, this would become difficult to negotiate. Even a purely technical Treaty amendment which only removes references to the UK would take some time to ratify, so one would expect work to get underway soon, if it is going to happen – although there could be a legal problem in that in principle the UK ought to participate in the Treaty amendment process as long as it is a member. One solution might be to draw up such a treaty after Brexit Day, and backdate its application date.

Paras 30 and 32 of the draft EP resolution refer to possible changes to EU law as a result of the UK’s departure as an issue for the EU-27, although they make no specific reference to Treaty amendment.

These guidelines define the framework for negotiations under Article 50 TEU and set out the overall positions and principles that the Union will pursue throughout the negotiation. The European Council will remain permanently seized of the matter, and will update these guidelines in the course of the negotiations as necessary.

Article 50(2) states that the negotiation must take place ‘in light of’ these guidelines. It is unclear precisely what legal effect this has. For instance, could an outvoted Member State challenge the Article 50 deal for breaching the guidelines? As it happens, there is a pending ECJ case which raises (among other things) the legal effect of European Council guidelines in the asylum context, discussed here.

The predominant effect of the guidelines will of course be political; note that the European Council is prepared to ‘update’ them (ie amend or elaborate upon them further). They will necessarily be fleshed out in detail in a rather longer treaty text, on the basis of drafts tabled and negotiated by each side.

I. Core principles

1. The European Council will continue to base itself on the principles set out in the statement of Heads of State or Government and of the Presidents of the European Council and the European Commission on 29 June 2016. It reiterates its wish to have the United Kingdom as a close partner in the future. It further reiterates that any agreement with the United Kingdom will have to be based on a balance of rights and obligations, and ensure a level-playing field. Preserving the integrity of the Single Market excludes participation based on a sector-by-sector approach. A non-member of the Union, that does not live up to the same obligations as a member, cannot have the same rights and enjoy the same benefits as a member. In this context, the European Council welcomes the recognition by the British Government that the four freedoms of the Single Market are indivisible and that there can be no "cherry picking".

The June 2016 statement can be found here. The second sentence matches the UK government’s objective of a future close partnership. Any ‘sectoral’ deals are excluded; this contradicts the UK government position, which had the objective of obtaining such deals. In any event sectoral deals would be potentially vulnerable to challenge under WTO rules, which require any free trade agreement to cover ‘substantially’ all trade in goods (Article XXIV GATT) or services (Article V GATS). It would however be possible for mutual recognition agreements to cover certain sectors of trade. There is no reference in the guidelines to the UK government’s objective to achieve some form of special customs agreement, except implicitly as regards Northern Ireland.

2. Negotiations under Article 50 TEU will be conducted as a single package. In accordance with the principle that nothing is agreed until everything is agreed, individual items cannot be settled separately. The Union will approach the negotiations with unified positions, and will engage with the United Kingdom exclusively through the channels set out in these guidelines and in the negotiating directives. So as not to undercut the position of the Union, there will be no separate negotiations between individual Member States and the United Kingdom on matters pertaining to the withdrawal of the United Kingdom from the Union.

It’s not clear whether the ‘single package’ refers more narrowly to the specific issues which the EU wants to negotiate first, or the talks on later status of UK/EU relations. This is important because there is a strong case for concluding a first treaty on withdrawal issues, including in particular the status of UK and EU citizens who have moved between Member States, separately from any subsequent treaties.

This paragraph reconfirms the ‘no separate negotiations’ point already made in the guidelines. As noted above, this does not prevent some informal bilateral discussions on specific issues. For instance, UK/Ireland discussions on border issues could be significant, and the guidelines in effect require separate talks between the UK and Spain regarding Gibraltar. They also refer to bilateral UK/Cyprus issues.

Article 50 refers to the EU acting by qualified majority vote (without the UK voting), so the common assumption that any Article 50 deal must be agreed unanimously and ratified by all national parliaments is incorrect. However, some aspects of the overall Brexit deal (ie on future relationship) might entail unanimous voting and national ratification; and the EU might be unwilling to proceed with the Article 50 deal against the opposition of one or more Member States.

II. A phased approach to negotiations

3. On the date of withdrawal, the Treaties will cease to apply to the United Kingdom, to those of its overseas countries and territories currently associated to the Union, and to territories for whose external relations the United Kingdom is responsible. The main purpose of the negotiations will be to ensure the United Kingdom's orderly withdrawal so as to reduce uncertainty and, to the extent possible, minimise disruption caused by this abrupt change.

This refers to the entities referred to in Article 353 TFEU, ie Gibraltar, the Channel Islands, the Isle of Man, and some small remaining semi-colonies. For a discussion of their status, see this House of Lords report. The reference to minimising disruption is important, although qualified (‘to the extent possible’).

To that effect, the first phase of negotiations will aim to:
settle the disentanglement of the United Kingdom from the Union and from all the rights and obligations the United Kingdom derives from commitments undertaken as Member State;
provide as much clarity and legal certainty as possible to citizens, businesses, stakeholders and international partners on the immediate effects of the United Kingdom's withdrawal from the Union.
The European Council will monitor progress closely and determine when sufficient progress has been achieved to allow negotiations to proceed to the next phase.

This asserts the EU’s determination to discuss withdrawal issues before trade issues, whereas the UK has wanted to discuss them in tandem. However, the EU guidelines do not require the withdrawal issues to be fully agreed before talks on post-Brexit issues get underway. The ‘sufficient progress’ test is deliberately flexible; it can be argued that agreement in principle on the main issues will be sufficient, even if some details have to be worked out.  Para 14 of the EP draft resolution largely matches this, although the EP will not have any formal role in determining whether ‘substantial progress’ (the slightly stricter test which the EP proposes) has been met.

4. While an agreement on a future relationship between the Union and the United Kingdom as such can only be concluded once the United Kingdom has become a third country, Article 50 TEU requires to take account of the framework for its future relationship with the Union in the arrangements for withdrawal. To this end, an overall understanding on the framework for the future relationship could be identified during a second phase of the negotiations under Article 50. The Union and its Member States stand ready to engage in preliminary and preparatory discussions to this end in the context of negotiations under Article 50 TEU, as soon as sufficient progress has been made in the first phase towards reaching a satisfactory agreement on the arrangements for an orderly withdrawal.

The EP draft resolution (point 15) also rules out conclusion of a ‘future relationship’ treaty before Brexit, so it seems unlikely to happen.

This discussion of a ‘future relationship’ between the UK and the EU (concerning trade and other issues) is separate from a possible ‘transitional arrangement’ referred to in the next paragraph (‘may also’). So it is unclear whether a transitional deal could also only be concluded after the UK leaves.

In any event, it is arguable that a longer-term deal could be concluded (or at least signed and applied provisionally) on the date of withdrawal, or failing that, it could be backdated to that date. Such an approach would be consistent with the EU’s objectives of an orderly withdrawal and minimising disruption. However, the references to ‘preliminary and preparatory’ and ‘overall understanding’ seem to rule that out.

In any event, it would be hard in practical terms to negotiate all the details of the long-term framework within the time available, even though this is Theresa May’s objective. This is particularly true given that some of the available time will be taken up negotiating withdrawal issues before moving on to the ‘future relationship’ talks.

5. To the extent necessary and legally possible, the negotiations may also seek to determine transitional arrangements which are in the interest of the Union and, as appropriate, to provide for bridges towards the foreseeable framework for the future relationship. Any such transitional arrangements must be clearly defined, limited in time, and subject to effective enforcement mechanisms. Should a time-limited prolongation of Union acquis be considered, this would require existing Union regulatory, budgetary, supervisory and enforcement instruments and structures to apply.

It is certainly in the EU’s interest to maintain trade flows and other relationships with the UK, and as noted above this would be consistent with the EU’s objectives of an orderly withdrawal and minimising disruption. They would obviously be ‘necessary’ to the end of avoiding a switch to a WTO-only framework for trade between the UK and the EU. Again as noted already, this paragraph is unclear about when a transitional deal might be concluded, although as compared to the previous paragraph it does not explicitly state that talks could not be completed or that an agreement could not be concluded before Brexit Day. It is also vague about the relevant legal base: is it Article 50 (qualified majority vote) or other Treaty provisions (more likely to entail unanimity and national ratification)? The exact scope of Article 50 is unclear legally; the wording here suggests some uncertainty about what is ‘legally possible’, which it may be necessary to ask the ECJ to clarify (see Annex II).

The transitional deal would have to be limited in time, which could reassure those who do not want the UK to remain in such an arrangement indefinitely. It is not expressly clear that all aspects of EU membership would be carried over. In particular, it is not clear that the only transition which the EU would consider would be continuing in force the EU acquis. The distinction is important because a) flexibility on this issue could involve ending the free movement of people, or continuing it with an ‘emergency brake’, from Brexit Day; and b) an ‘enforcement mechanism’ might fall short of continuing EU ‘existing regulatory, budgetary, supervisory and enforcement’ measures. The latter phrase would logically entail keeping in force the ECJ’s current jurisdiction, while a mere ‘enforcement mechanism’ could entail a more limited role for the ECJ (no references from UK courts or direct impact on UK law) or a different dispute mechanism system entirely, such as using the EFTA Court which already exists and rules on certain EU law issues as regards Norway, Iceland and Liechtenstein. However, the more different a transitional system would be from current EU membership, the longer it would take to negotiate, and the bigger the risk of running out of time.

The UK government’s position seems to contemplate some form of interim transitional period, without calling it such. The Chatham House speech refers instead to ‘phased implementation’, during which aspects of EU law will still apply. But despite the difference in wording, fundamentally there are strong similarities between the UK and EU positions here.

The EP draft resolution is similar (para 28). It sets a maximum time limit of three years for the transitional deal, although otherwise it is as vague as the European Council draft guidelines.

6. The core principles set out above should apply equally to the negotiations on an orderly withdrawal, to any preliminary and preparatory discussions on the framework for a future relationship, and to any form of transitional arrangements.

7. The two year timeframe set out in Article 50 TEU ends on 29 March 2019.

It should be noted that Article 50(3) provides that the withdrawal agreement may set a different date (which could implicitly be later or earlier), and that the UK and the remaining EU, voting unanimously, may postpone the overall deadline.

III. Agreement on arrangements for an orderly withdrawal

8. The right for every EU citizen, and of his or her family members, to live, to work or to study in any EU Member State is a fundamental aspect of the European Union. Along with other rights provided under EU law, it has shaped the lives and choices of millions of people. Agreeing reciprocal guarantees to settle the status and situations at the date of withdrawal of EU and UK citizens, and their families, affected by the United Kingdom's withdrawal from the Union will be a matter of priority for the negotiations. Such guarantees must be enforceable and non-discriminatory.

The EU, like the UK, places priority on an early deal on the rights of each other’s citizens. While some had claimed that the EU could not or would not address this issue in the Article 50 talks, it clearly intends to do so. The reference to ‘non-discriminatory’ arrangements confirms that there will not be bilateral talks on this issue, as some people had expected. While some have claimed that the EU lacks legal power to regulate the position of non-EU citizens, Articles 77-79 TFEU confer such powers and the EU has adopted a number of laws in this area. In any event, it is strongly arguable that Article 50 confers power upon the EU to negotiate any issue which arose pursuant to the withdrawing Member State’s membership of the European Union, and the status of UK and EU citizens is one obvious example of such an issue.

The reference to an ‘enforceable’ guarantee does not necessarily entail using the ECJ. There could be some other form of dispute settlement, or a commitment to make the agreement binding in national law and to hold discussions about any issues which arise might suffice.

It should be noted that the guidelines make no reference to the idea, promoted by some in the EP, of an optional right for UK citizens to retain EU citizenship. The EP draft resolution does refer to this issue, but states that such arrangements should be reciprocal (para 27). It is hard to imagine the UK government agreeing to such a deal.

9. Also, the United Kingdom leaving the Union will impact EU businesses trading with and operating in the United Kingdom and UK businesses trading with and operating in the Union. Similarly, it may affect those who have entered into contracts and business arrangements or take part in EU-funded programmes based on the assumption of continued British EU membership. Negotiations should seek to prevent a legal vacuum once the Treaties cease to apply to the United Kingdom and, to the extent possible, address uncertainties.

This indicates a general intention to retain contracts and legal arrangements in force, if they have been concluded before Brexit Day. This could be relevant to research funding, regional funding or farm subsidies, for instance. It could also be the basis for arguing that UK banks who already have a licence to sell financial services to the EU market can retain it.

10. A single financial settlement should ensure that the Union and the United Kingdom both respect the obligations undertaken before the date of withdrawal. The settlement should cover all legal and budgetary commitments as well as liabilities, including contingent liabilities.

This is likely to be one of the most difficult issues to negotiate. The draft guidelines do not put a specific figure on the ‘bill’, although press reports note amounts such as €50 billion. There is no reason why the amount of any bill should be paid upfront, as some of it relates to funding over the next few years and to pensions payable over the longer term. For detailed discussions of this issue see this House of Lords report and this Brueghel report.

11. The Union has consistently supported the goal of peace and reconciliation enshrined in the Good Friday Agreement, and continuing to support and protect the achievements, benefits and commitments of the Peace Process will remain of paramount importance. In view of the unique circumstances on the island of Ireland, flexible and imaginative solutions will be required, including with the aim of avoiding a hard border, while respecting the integrity of the Union legal order. In this context, the Union should also recognise existing bilateral agreements and arrangements between the United Kingdom and Ireland which are compatible with EU law.

The EU and UK both make a priority of addressing the border issue between Ireland and Northern Ireland. (It should be noted that there is no specific reference to Scotland in the EU guidelines; the EP resolution merely notes that Scotland voted to Remain). There are indications of great willingness to compromise here (‘flexible and imaginative solutions’) and it should be noted that Protocols to the Treaties require the EU to facilitate the operation of the Common Travel Area between the UK and Ireland (the words ‘should also recognise’ these arrangements do not adequately take account of this legal obligation). However, the Protocols do not lay down in detail what happens in the event of UK withdrawal, which therefore has to be the subject of negotiations. Equally, arguments about the history of the UK and Ireland are irrelevant, as there has never been a previous situation when one of those countries was in the EEC/EC/EU and the other was not.

12. The Union should agree with the United Kingdom on arrangements as regards the Sovereign Base Areas of the United Kingdom in Cyprus and recognise in that respect bilateral agreements and arrangements between the Republic of Cyprus and the United Kingdom which are compatible with EU law, in particular as regards the situation of those EU citizens resident or working in the Sovereign Base Areas.

There is a specific Protocol to the 2003 Accession Treaty on the position of the UK sovereign base in Cyprus, which will presumably have to be amended in order to take account of any such agreement. Here, despite the overall EU-wide approach to talks, there is recognition that there will be a bilateral agreement between the UK and Cyprus alongside an EU-UK agreement.

13. Following the withdrawal, the Union with 27 Member States will continue to have the rights and obligations of the Union with 28 Member States in relation to international agreements. The United Kingdom will no longer be covered by agreements concluded by the Union or by Member States acting on its behalf or by both acting jointly. The European Council expects the United Kingdom to honour its share of international commitments contracted in the context of its EU membership. In such instances, a constructive dialogue with the United Kingdom on a possible common approach towards third country partners and international organisations concerned should be engaged.

The EU position is that the UK ceases to be part of any treaty with non-EU countries within the scope of EU law – whether concluded by the EU alone, by the Member States alone (but affecting EU law) or by both together. This is particularly relevant to trade agreements (see discussion by Markus Gehring here) but affects other agreements as well (for example, environmental deals).

The UK will likely seek to secure replacement agreements with the countries concerned. This is even encouraged by the EU, which seeks a ‘constructive dialogue’ on these issues. The reference to ‘international organisations’ most obviously refers to the WTO.

In practical terms, the issue which particularly arises is ‘tariff rate quotas’, ie allowing in a certain amount of products at a low tariff. Say the EU allows 100,000 tons of olives imported from Morocco at a low tariff: the obvious solution is to split that between the UK and the remaining EU based on recent trade flows (ie how much of those olives were imported into the UK in the last three years, as compared to the rest of the EU?).

Non-EU countries will obviously have to agree to this process, and the UK might want to focus particularly on replicating those agreements with a major impact on UK exports: for instance, the EU agreements on the protection of the name ‘Scotch Whisky’.

14. While the future location of the seats of EU agencies and facilities located in the United Kingdom is a matter for the 27 Member States, arrangements should be found to facilitate their transfer.

The intention is obviously to move these bodies as soon as possible after Brexit Day, if not before. The UK will not be involved in decisions on where they move to, but will be involved in the logistics of moving them.

15. Arrangements ensuring legal certainty and equal treatment should be found for all court procedures pending before the Court of Justice of the European Union upon the date of withdrawal that involve the United Kingdom or natural or legal persons in the United Kingdom. The Court of Justice of the European Union should remain competent to adjudicate in these procedures. Similarly, arrangements should be found for administrative procedures pending before the European Commission and Union agencies upon the date of the withdrawal that involve the United Kingdom or natural or legal persons in the United Kingdom. In addition, arrangements should be foreseen for the possibility of administrative or court proceedings to be initiated post-exit for facts that have occurred before the withdrawal date.

The EU assumes that EU court cases pending on Brexit Day should remain within the competence of the Court to decide. This is a classic transitional issue and the EU suggestion makes sense, since the cases concern the pre-Brexit legal position. (Compare to the planned Great Repeal Bill, which will provide that pre-Brexit ECJ case law will still be binding). It is slightly more ambiguous as regards pending administrative procedures like competition law and state aid (‘arrangements should be found’). The same applies to cases with EU law issues pending in the UK courts on Brexit Day, or which are brought in the UK courts in summer 2019 relating to a 2018 tax bill, for instance.

This paragraph is too limited, as the transitional deal needs to take explicit account of all legal proceedings pending on Brexit Day pursuant to EU law, not just those which involve or may involve the EU institutions: for instance a pending claim to recognise a German court judgment in the UK, or the French authorities’ obligation to execute a European Arrest Warrant issued by the UK before that date. Most implementation of EU law is carried out by national courts and administrations, not EU bodies, and the transitional rules should take account of this.

16. The withdrawal agreement should include appropriate dispute settlement mechanisms regarding the application and interpretation of the withdrawal agreement, as well as duly circumscribed institutional arrangements allowing for the adoption of measures necessary to deal with situations not foreseen in the withdrawal agreement. This should be done bearing in mind the Union's interest to effectively protect its autonomy and its legal order, including the role of the Court of Justice of the European Union.

The institutional arrangements will presumably entail an EU/UK Joint Committee with the power to take decisions by joint agreement. There are similar bodies in other EU treaties with non-EU states. The first and third sentences, read together, do not explicitly insist that the ECJ have jurisdiction over the withdrawal agreement (although it will inevitably have jurisdiction as regards the EU side). The reference to ensuring EU autonomy reflects ECJ case law which states that treaties with non-EU countries cannot affect the separate development of EU law or the essential features of the Court’s powers: see Opinion 1/91 and Opinion 1/00. Compare with point 17 of the EP resolution, which explicitly calls for the ECJ to have jurisdiction over the withdrawal agreement.

IV. Preliminary and preparatory discussions on a framework for the Union - United Kingdom future relationship

17. The European Council welcomes and shares the United Kingdom's desire to establish a close partnership between the Union and the United Kingdom after its departure. While a relationship between the Union and a non Member State cannot offer the same benefits as Union membership, strong and constructive ties will remain in both sides' interest and should encompass more than just trade.

The EU accepts in general the UK position of having a close partnership, including but going beyond trade. There is no explicit reference to the form of the relationship, which is relevant given that it could affect whether the EU side has to vote unanimously and ask national parliaments to ratify any treaty. (EU treaties with non-EU states can be partly applied provisionally pending such ratification).Point 22 of the EP draft resolution hints at a possible association agreement: this entails unanimous voting and usually also national ratification.

18. The British government has indicated that it will not seek to remain in the single market, but would like to pursue an ambitious free trade agreement with the European Union. Based on the Union's interests, the European Council stands ready to initiate work towards such an agreement, to be finalised and concluded once the United Kingdom is no longer a Member State.

The EU accepts the UK position of seeking a far-reaching free trade deal, rather than continued single market participation. Note that there is no reference to continuing with the free movement of persons or contributions to the EU budget – two key objectives of the UK side. The timing is an issue, as noted already: no finalisation or conclusion until after Brexit Day, which means that a transitional deal will be important in the meantime. Although in theory this could be on or soon after Brexit Day, there is unlikely to be enough time for that.

19. Any free trade agreement should be balanced, ambitious and wide-ranging. It cannot, however, amount to participation in the Single Market or parts thereof, as this would undermine its integrity and proper functioning. It must ensure a level playing field in terms of competition and state aid, and must encompass safeguards against unfair competitive advantages through, inter alia, fiscal, social and environmental dumping.

It is common for EU trade agreements to make some reference to state aid and competition law, although only some of them require the non-EU state to apply EU competition and state aid law as such. The guidelines leave it open as to what exactly the EU side will aim for. In any event WTO rules also contain some restrictions on granting subsidies and set out the possibility for trading partners to retaliate if subsidies are granted, although these rules are less far-reaching and enforceable than those applying to EU Member States. (Note that the EU does not ban state aids entirely, but restricts them to certain cases).

The ‘safeguards against unfair competitive advantages’ the EU side wants refers to – among other things – ‘fiscal, social and environmental dumping’. This does not explicitly refer to the adoption of EU law by the UK, and in any event there is no EU law on minimum corporate tax rates. (Compare to para 24 of the EP resolution, which refers more explicitly to EU laws, although to ‘tax evasion and avoidance’ as distinct from tax rates). The UK might argue that it would be sufficient to remain party to international treaties on environmental law and social protection (in the ILO, the Council of Europe and the UN Covenant on Economic, Social and Cultural Rights), and to hold regular discussions on corporate tax rates – which could cut both ways. In any event, many in the UK would welcome safeguards against cuts in environmental and social standards, and would be concerned about how public services could be funded in the event of large cuts in corporate tax.

It remains to be seen what exactly the EU side would accept as safeguards, but the EU’s position should be seen in its overall context: as noted already, the EU is not insisting on free movement of people or financial contributions. It is striking that the EU side makes no explicit references to services, where the UK has a trade surplus. If the UK wants to maintain that strong surplus by having an advanced trade relationship that offers more free trade in services than the EU usually agrees (but still falling short of single market participation) this is the condition which the EU wants to set.

20. Beyond trade, the EU stands ready to consider establishing a partnership in other areas, in particular the fight against terrorism and international crime as well as security and defence.

This matches the UK’s position. Neither side sets out any real details here.

21. The future partnership must include appropriate enforcement and dispute settlement mechanisms that do not affect the Union's autonomy, in particular its decision-making procedures.

There is no specific mention of the ECJ here, and the autonomy issue was discussed above. It’s not usual for the Court to have jurisdiction as regards non-EU states, bar a few exceptions like the European Aviation Area treaty, which facilitates aviation between EU and other European countries. But it is common for the EU to agree dispute settlement similar to the WTO dispute settlement system in agreements with non-EU countries. Interestingly, the EU does not use such systems in its trade agreements in practice, although it does often use the WTO.

In short, the WTO system provides for panels of experts to decide on whether there is a breach of WTO law; their decisions can be appealed to an Appellate Body. If a WTO party which was found to have breached WTO law does not comply with these rulings, the victorious party which brought the complaint can be authorised to retaliate against it with proportionate trade sanctions. Dispute settlement bodies are not unique to the EU and WTO – there is an active system in the NAFTA agreement between the US, Canada and Mexico, for instance. Such systems fall short of the legal effect of EU law in national legal systems, but still place some constraints upon the parties to trade treaties.

22. After the United Kingdom leaves the Union, no agreement between the EU and the United Kingdom may apply to the territory of Gibraltar without the agreement between the Kingdom of Spain and the United Kingdom.

This clause has attracted much overreaction. It is not a claim to Gibraltar territory or joint sovereignty, so there is no need to respond to it with talk of military action. It merely acknowledges that the EU will not apply post-Brexit treaties to the UK unless the UK and Spain have separately agreed to this. It will be up to the UK and Spain to find agreement for each treaty, or failing that to accept that the treaty in question will either not be concluded or not apply to Gibraltar.

V. Principle of sincere cooperation

23. Until it leaves the Union, the United Kingdom remains a full Member of the European Union, subject to all rights and obligations set out in the Treaties and under EU law, including the principle of sincere cooperation.

This reflects the UK’s continued position that it will apply EU law until Brexit Day. Point 5 of the EP draft resolution matches it.

24. The European Council recognises the need, in the international context, to take into account the specificities of the United Kingdom as a withdrawing Member State, provided it remains loyal to the Union's interests while still a Member. Similarly the Union expects the United Kingdom to recognise the need of the 27 Member States to meet and discuss matters related to the situation after the withdrawal of the United Kingdom.

The first sentence is ambiguous about a key issue: can the UK already discuss trade deals with non-EU countries? It can’t conclude them without violating EU law (para 23); normally it could not negotiate them either, but does this sentence accept the argument that discussions on a post-Brexit deal would be acceptable, as a consequence of the UK’s position as a withdrawing Member State? The second sentence asserts the remaining EU’s right to meet without the UK, presumably going beyond the talks relating to Brexit without the UK present as referred to in Article 50. However, such meetings must remain informal, as the next paragraph confirms.  

Compare to point 6 of the draft EP resolution, which more explicitly argues that the UK cannot negotiate with non-EU countries before Brexit, and argues that the UK should be excluded from EU trade talks with non-EU countries if it does. Such an exclusion would not be legal; the remedy in such cases of alleged breach of EU law is for the Commission or another Member State to bring the UK to the ECJ.  

25. While the United Kingdom is still a member, all ongoing EU business must continue to proceed as smoothly as possible at 28. The European Council remains committed to drive forward with ambition the priorities the Union has set itself. Negotiations with the United Kingdom will be kept separate from ongoing Union business, and shall not interfere with its progress.

The UK retains its formal position as a Member State until Brexit Day, although obviously it will have declining influence as there will be little interest in addressing its concerns and the other Member States merely have to wait out any veto or participation in a blocking minority vote by the UK.

VI. Procedural arrangements for negotiations under Article 50

The European Council endorses the arrangements set out in the statement of 27 Heads of State or Government on 15 December 2016.

These procedural arrangements are discussed in the next annex.


Annex II – Procedural rules

Here are the December 2016 procedural rules, with annotations.

1. The first step following the notification by the United Kingdom will be the adoption by the European Council of guidelines that will define the framework for negotiations under Article 50 TEU and set out the overall positions and principles that the EU will pursue throughout the negotiation. The European Council will remain permanently seized of the matter, and will update these guidelines in the course of the negotiations as necessary.

This refers to the guidelines of the European Council (the Member States’ Heads of State and Government), the draft of which is now available and was discussed above. Article 50 requires the guidelines to be adopted at the start of the process.

2. After the adoption of the guidelines, the European Council will invite the General Affairs Council to proceed swiftly with the adoption of the decision authorising the opening of the negotiations, following a recommendation by the European Commission, and to deal with the subsequent steps in the process. The Council will also adopt negotiating directives on substance as well as on the detailed arrangements governing the relationship between the Council and its preparatory bodies on the one hand and the Union negotiator on the other. These negotiating directives may be amended and supplemented as necessary throughout the negotiations, to reflect the European Council guidelines as they evolve.

This follows the normal process of EU treaty negotiation with non-EU countries, as set out in Article 218 TFEU: a Commission recommendation to start talks, and a Council mandate with detailed negotiation objectives, which may be amendment throughout the talks. That mandate will likely be more detailed than these guidelines. The General Affairs Council is made up of Member States’ foreign or Europe ministers, and meets monthly. The Council’s ‘preparatory bodies’ consist of working groups and Member States’ permanent representatives (the latter meets weekly, and is known as ‘Coreper’).

3. The Council will be invited to nominate the European Commission as the Union negotiator. The Commission's nomination of Michel Barnier as chief negotiator is welcome. To ensure transparency and build trust, the Union negotiator's team will be ready to integrate a representative of the rotating Presidency of the Council. Representatives of the President of the European Council will be present and participate, in a supporting role, in all negotiation sessions, alongside the European Commission representatives. The Union negotiator will systematically report to the European Council, the Council and its preparatory bodies.

Article 50 leaves open who the negotiators will be, and some had assumed it would be the Council. Nominating the Commission follows the usual approach under Article 218 TFEU. However, including a representative of the Council Presidency (which rotates every six months) and the European Council President (Donald Tusk) in the talks is an innovation, which reflects their importance. The obligation to report back to EU bodies reflects Article 218 TFEU.

4. Between the meetings of the European Council, the Council and Coreper, assisted by a dedicated Working Party with a permanent chair, will ensure that the negotiations are conducted in line with the European Council guidelines and the Council negotiating directives, and provide guidance to the Union negotiator.

Article 218 TFEU (and Article 207 TFEU, regarding trade) refer to Council working parties supervising Commission negotiators, so this is nothing new. As regards trade, the Commission refers to the committee which supervises it as the ‘mothers-in-law’ (don’t shoot the messenger!).

5. The members of the European Council, the Council and its preparatory bodies representing the United Kingdom will not participate in the discussions or in the decisions concerning it.

This simply respects the wording of Article 50, extended (logically enough) to the Council’s working parties, not just to the official meetings of ministers or heads of state.

6. Representatives of the 27 Heads of State or Government (Sherpas/Permanent Representatives) will be involved in the preparation of the European Council as necessary. Representatives of the European Parliament will be invited at such preparatory meetings.

This gives a slightly greater role than usual to the European Parliament.

7. The Union negotiator will be invited to keep the European Parliament closely and regularly informed throughout the negotiation. The Presidency of the Council will be prepared to inform and exchange views with the European Parliament before and after each meeting of the General Affairs Council. The President of the European Parliament will be invited to be heard at the beginning of meetings of the European Council.

The first sentence reflects Article 218 TFEU, which calls for the EP to be fully and immediately informed throughout negotiations. The second and third sentences go into more detail, and give a bigger role to the EP than usual during EU talks with non-EU countries. As noted already, however, the EP is not a negotiator as such, although its power to give its consent to the final deal (not referred to explicitly here) nevertheless means it is a significant actor. The additional meetings referred to here will be a forum for the EP to influence the negotiations.

There is no reference to the ECJ, which can rule on legal issues concerning draft treaties with non-EU countries pursuant to Article 218 TFEU, at the request of the Commission, Council, EP or a Member State. It is not clear whether the whole of Article 218 necessarily applies to the Brexit treaty process, since Article 50 only refers to some of it. There are also other ways that Brexit legal questions can reach the ECJ: national courts can ask the Court about them, the Commission can sue a Member State (or Member States can sue each other), or Member States or the EU institutions can sue the [other] EU institutions for various aspects of their conduct of the negotiation.

Barnard & Peers: chapter 27

Photo credit: the Register