“We must not make a scarecrow of the law”
(Measure for Measure, 2.1.1)
Charlotte O’Brien, Senior Lecturer, York Law School, University of York
Laurent Pech, Professor of European Law, Jean Monnet Chair of EU Public Law at Middlesex University London
This post primarily aims to clarify the scope of EU free movement rules with the view of addressing some of the most common (legal) misconceptions one may regularly encounter in the British media and elsewhere. We argue that the much of the criticism of free movement has targeted a ‘straw man’, with little regard to the actual regime.
In the UK, for instance, there seems to be a wide consensus across the political spectrum on the need to restrict ‘current EU freedom of movement rules’ yet ‘current’ EU rules are often misunderstood or misrepresented. To argue that the UK has simply ‘no control’ over immigration from the rest of the EU is inaccurate. Similarly, one may regret that some EU actors tend to paint freedom of movement as one which could not be subject to restrictions and one which would be inherently connected with rules on the free movement of goods, capital and services.
These aspects and many others will be addressed below via 10 questions & answers, which should be of interest to members of the general public looking for an (hopefully) accessible and legal overview of EU free movement law. The post will end with a brief outline of the main models and possible options for the UK to consider when it will have to agree the ‘framework for its future relationship with the Union’ to quote Article 50 of the Treaty on European Union (hereinafter: TEU).
(1) Does EU Law guarantee an absolute right to move and reside anywhere in the EU?
No. Treaty rights of free movement are subject to limitations.
Article 45(3) of the Treaty on the Functioning of the European Union (hereinafter: TFEU) states that the rights to (a) accept offers of employment, (b) move freely between States to take up employment, (c) reside in other Member States, and (d) the right to stay in another Member State after employment has finished, are subject to ‘limitations justified on grounds of public policy, public security or public health’. This means, to simplify, that national authorities can adopt restrictive measures on a case-by-case basis against EU workers on any of these three grounds. Arguably the most mutable of these grounds is the public policy limitation, and Gareth Davies has argued that this limitation has been underexplored when it comes to free movement concerns, suggesting that greater use of permitted restrictions might have avoided free movement becoming as contentious an issue as it did in the UK.
Article 45(4) TFEU adds a further exception – that the free movement provisions shall not apply to employment in the public service, so that where roles are thought to be nationally sensitive, involving participation in the exercise of powers conferred by public law and duties designed to safeguard the general interests of the State, they may be reserved for nationals. This does not amount to a right to employ non-nationals and then treat them differently – once admitted to the labour market, they should not be discriminated against.
Similarly, there is another derogation in respect of activities connected with the exercise of official authority for self-employed persons and providers of services (see Articles 51 and 62 TFEU). Discrimination against non-nationals is allowed but only with regard to positions involving a direct and specific connexion with the exercise of official authority. This is why it would not for instance be lawful to adopt a general ban on non-UK academics providing‘advisory work’ for UK public bodies such as the Foreign and Commonwealth Office on Brexit related matters.
The key point remains that EU Treaties have always clearly provided that free movement rights are not absolute and included in addition a number of permanent derogations to the general principle of non-discrimination on grounds of nationality. Unsurprisingly, however, a number of guiding principles have been developed by the Court of Justice to prevent an abusive reliance on these exceptions by national authorities.
Secondary EU legislation forms the ‘specific expression’ of the EU Treaty free movement rights, and sets out additional and more specific conditions and limitations. The key condition for those concerned about benefit tourism is that an EU national does not have a right to reside in a host Member State if they become an “unreasonable burden” upon the public purse. Directive 2004/38 (the “EU citizens’ Directive”, which is the source of the main legal rules on this issue) sets out the ‘categories’ of people with an EU-law based right to reside in a host Member State for longer than three months – essentially, workers/the self-employed and their family members, students, and the self sufficient. Students must have health insurance and make a declaration of sufficient resources.
In recent years, CJEU rulings have made some of these limitations more explicit. The Dano case made clear that someone who moved for the ‘sole’ purpose of claiming benefits would not have a right to reside in a member state under EU law. The Breycase approved of member states setting a right to reside condition in order for a claimant to be eligible for particular social security benefits, not just social assistance benefits. So member states have discretion over the arrangement of their welfare systems and eligibility conditions, and are entitled to set conditions tied to having a right to reside that only apply to EU nationals, and make it impossible for economically inactive EU nationals to claim benefits.
And in Alimanovicand Garcia Nieto, the CJEU emphasised that EU nationals do not have entitlement to social assistance benefits during their first three months of residence in a host Member State, or at any point at which they are classed as a ‘jobseeker’, even where the benefit in question appears to be an unemployment benefit. EU jobseekers are entitled to claim benefits that facilitate access to the labour market, but not where the benefit has a social assistance element and primary aim of the benefit is the preservation of dignity. In the Commission v UK case the CJEU appear to have extended the exclusion to full social security benefits, not just those with a social assistance element. Economically inactive EU nationals must be self sufficient in order to have a right to reside, and EU nationals who are workers must demonstrate that their work is genuine and effective.
Furthermore, there are some geographic limitations – free movement of persons applies for people moving from the Overseas Countries or Territories associated with the Union (OCTs), to the EU, but does not apply to people moving from the EU to OCTs as will be explained below.
(2) Does EU Law prevent temporal and geographical exceptions?
It would be wrong to think that EU law has never allowed for some extensive geographical and temporal exceptions with respect to the free movement of people.
While the default principle is that EU law applies in the territory of all the Member states, some exceptions exist especially when it comes to overseas regions and territories of relevant Member States (so much so that there is a specialised field of study known as EU Law of the Overseas). This means for instance EU free movement law does not uniformly apply to the whole of the territories of all the Member States. The relevant legal framework is rather complex. In a nutshell, there is free movement in one direction only. Inhabitants of the countries making up the Netherlands Antilles may move to any EU Member State as they are Dutch and EU citizens, but EU citizens from any other Member State do not have symmetrical rights to move there. This is lawful provided that the authorities of the countries of the Dutch Caribbean apply the same restrictions on entry and residence equally to the nationals of the Member State with which it is associated and the nationals of other EU Member States (authors are grateful for Prof Kochenov’s insight on this issue. For more details on ‘one-way EU free movement’ with respect to the OCTs, see his article here).
But one does not have to go overseas to see territories where the application of EU law may vary dramatically depending on the subject matter. To give a single example, while EU rules on free movement of goods apply to the Channel Islands, the rules relating to the free movement of persons and services do not (see Protocol No 3 to the UK’s Act of Accession).
To make matters even more complex, EU free movement rules may apply to nationals of non-EU Member States. For instance, they apply to Swiss nationals and nationals of Norway, Iceland and Liechtenstein (though note that free movement is in effect asymmetrical as far as Liechtenstein is concerned – see question (7) below) which are not EU Member States, by virtue – to simplify – of a number of bilateral agreements between the EU and those countries.
With respect to temporal exceptions to the free movement of people within the EU, one may for instance refer to the Treaties of accession of 2003, 2005 and 2012, which included provisions allowing for transitional restrictions to be imposed by existing EU Member States on workers from the new EU Member States that joined the EU in 2004, 2007 and 2013 respectively. To give a single example, one may cite the chapter entitled ‘freedom of movement for persons’ of Annex XII to the 2003 Act of Accession regarding Poland. According to this text, the rules governing freedom of movement of workers and freedom to provide services shall apply subject to a number of transitional provisions on the basis of which the then EU-15 could adopt national measures in order to regulate ‘access to their labour markets by Polish nationals’ or adopt measures derogating from the freedom to provide services as far as companies established in Poland are concerned.
By contrast to most EU Member States, the UK decided not to derogate from the free movement of workers during the seven-year period with respect to nationals of the countries that joined the EU in 2004. In other words, the UK along with Ireland and Sweden sovereignly decided to open their labour markets directly from 1 May 2004 with respect to workers from countries such as Poland. The UK did however impose transitional provisions departing from the principle of equal treatment (see Question (3) below).
The UK also subsequently decided to derogate from free movement with respect to Bulgaria and Romania when these two countries joined the EU in 2007. This meant that workers from the two countries were not allowed, as a matter of principle, to work freely in the UK before 1 January 2013, that is, seven years after Bulgaria and Romania had joined the EU. To oversimplify, Romanian and Bulgarian nationals wishing to work in the UK had instead to secure the permission of the Home Office for doing so. The same has applied to Croatian workers since Croatia joined the EU in 2013.
(3) Has EU Law led to an uncontrolled migration regime within the EU?
The free movement system means that EU nationals and their family members who fulfil the conditions within EU law can move and reside freely within the EU, which means that Member States cannot impose quotas, or add additional conditions through an immigration points system – unless some derogation from Article 45 TFEU were agreed to be necessary on pressing grounds of, e.g. public policy.
Fulfilling the conditions set by the EU, in order to exercise free movement rights, typically means being a worker. This is therefore not uncontrolled migration, but the means of control is not immigration law – it is the labour market, which is the key to a right to reside and to stay in another Member State. This is an important point too often ignored. Moving away from this labour market based regulated form of intra-EU migration control would necessarily mean implementing instead a dirigiste bureaucratic form of migration control. This would not be a cost-neutral policy decision. It would likely be more time consuming and costly than the current market-based regulation system derived from EU law. One well informed expert suggested in this respect that ending the free movement of people in the UK would create a ‘bureaucratic nightmare’.
As noted above, Member States have in recent years been permitted to moderate the effect of accession of new states upon their labour markets by temporarily derogating from the free movement of workers with regard to new acceding states. While, as previously mentioned, the UK chose not to derogate from free movement for the A8 enlargement in 2004, it imposed a ‘worker registration scheme’ according to which A8 workers were not entitled to any out of work benefits until they had completed 12 months of registered work, and if the registration was not completed correctly, all subsequent work, until the lapse of the transition measures, was to considered unlawful, not conferring a right to reside or counting towards permanent residence. The transition measures ended in 2011.
The UK’s derogation from the free movement provisions in 2007 for the A2 States, Bulgaria and Romania, involved adopting a worker authorisation scheme, which essentially required Bulgarian and Romanian nationals to apply for authorisation to work on the basis of falling into an authorised category of work, ie specific sectors. These transition measures ended on the 1 January 2014. The UK also derogated from the free movement provisions on the accession of Croatia in 2013, and now requires Croatian nationals to obtain authorisation to work, and limits such authorisation to skilled work.
There was some implication in the lead up to the UK referendum that membership of the EU increased the ‘risk’ of unauthorised immigration and/or asylum seeking from third countries. But EU free movement law does not affect Member States’ (non-asylum) immigration regimes applied to non-EU nationals who are not family members of EU nationals. The Common European Asylum System provides a means for determining the state responsible for processing asylum claims, and under the problematic Dublin system this usually means that the country of first entry is responsible – putting most pressure on countries around the periphery of the EU, not the UK. A proposal for a reformed system for asylum burden sharing is in preparation, but the UK has secured an opt-out.
(4) Does EU Law prevent Member States from dealing with welfare tourism?
The short answer is a clear no. The CJEU has found that Member States are entitled to reserve the right to equal treatment as regards welfare benefits to those with an EU law-based right to reside.
To establish such a right, EU nationals must show that they fit one of the given categories in Directive 2004/38 – i.e. that they are workers, family members of workers self-sufficient, or students (who must declare self-sufficiency). EU law does not provide a right to reside for persons who move solely to claim benefits, and creates only limited rights for jobseekers, who are not entitled to claim social assistance.
There is no entitlement for those who move to seek work to social assistance for the first three months of residence; instead those who have been working elsewhere and who are entitled to a contributory job seeking benefit in their home State are entitled to bring it with them when they arrive in a host State, if they meet the conditions for ‘exporting’ their benefit.
As far as workers are concerned, EU law requires EU national workers to be treated equally with own state nationals for the purposes of social and tax advantages. In some cases this means that social security (but not social assistance) benefits can be exported to another Member State, so long as the EU national is a worker in the paying State. EU nationals who do not fall into these categories do not have a right to reside under EU law and so do not have equal access to welfare benefits.
Member States thus have some considerable freedom to limit benefit entitlement to those considered to be contributors, and so to avoid awarding benefits to those perceived to be benefit tourists. It is therefore misleading to suggest, as Theresa May did in August 2015, that EU law would guarantee a freedom to claim benefits.
However, in any discussion of the need to ‘deal’ with welfare tourism, we cannot ignore the absence of evidence that the problem exists. If we look to the take empirical evidence on the costs/benefits of EU immigration, e.g. the cost/benefit study by UCL, this suggests that EU nationals are net contributors, and ONS figures show that in the UK, EU nationals are less likely than UK nationals to be unemployed. The Department for Work and Pensions responded to a EU Commission query in 2013saying that they had no evidence of benefit tourism. Oxford University’s Migration Observatoryconcluded that the more recent the migrant’s arrival, the more likely the positive contribution. One may finally mention a report by the Centre for Economic Performance (CEP) at the London School of Economics, published on 11 May 2016, in which its authors argue that ‘the empirical evidence shows that EU immigration has not had significantly negative effects on average employment, wages, inequality or public services at the local level for the UK-born … At the national level, falls in EU immigration are likely to lead to lower living standards for the UK-born. This is partly because immigrants help to reduce the deficit: they are more likely to work and pay tax and less likely to use public services as they are younger and better educated than the UK-born. It is also partly due to the positive effects of EU immigrants on productivity.’
(5) Does EU Law prevent Member States to prevent abuse and fraud such as marriages of convenience?
EU free movement law is regularly presented as preventing national authorities from ‘controlling’ EU citizens seeking to move and reside in the UK. One must however emphasise that EU law cannot be relied upon in case of abuse. In other words, EU law explicitly entitles Member States to ‘adopt the necessary measures to refuse, terminate or withdraw any right conferred by [EU free movement law] in the case of abuse of rights or fraud’ and gives the example of marriages of convenience (see Article 35 of Directive 2004/38).
To prevent any abusive use of this ‘abuse clause’, whereby member states might unduly strip people of their rights, EU law also unsurprisingly provides that any measure adopted by a national authority on this basis must however be proportionate and subject to the usual procedural safeguards such as access to judicial and administrative redress procedures in the host Member State.
It would be wrong therefore to claim that EU law prevents Member States from tackling abusive reliance on EU free movement rights, which has been defined as any ‘artificial conduct entered into solelywith the purpose of obtaining the right of free movement and residence under [EU] law which, albeit formally observing of the conditions laid down by [EU] rules, does not comply with the purpose of those rules.’ Fraud, that is, any deliberate deception or contrivance made to obtain the EU right of free movement and residence may similarly be sanctioned by the relevant host Member State.
In practice, it is for each Member State to decide how best to tackle fraud and abuse of EU free movement rights. As noted above, EU law does not prevent Member States from investigating individual cases where there is a well-founded suspicion of abuse and adopt necessary sanctions in cases of proved abuse. It is not unusual for national authorities to choose not to go to the expense of doing so; a recent analysis on the Free Movement blog noted that ‘the latest figures suggest there are very few investigations and the Home Office is unwilling to release information on the outcome of the investigations.’ However, this lack of action is sometimes attributed to the legal constraints of EU free movement law (see e.g. the letter sent by David Cameron to Donald Tusk on 10 November 2015 citing inter alia ‘sham marriage’ to justify a rewriting of EU free movement rules). These legal constraints however merely call for a case-by-case assessment of any possible abuse of EU free movement rights and for any individual investigation to be carried out in accordance with fundamental rights. In other words, EU law only requires Member States to comply with the rule of law in this area.
(6) Does EU Law prevent Member States from removing criminals from their territories?
Some British newspapers regularly seek to paint the EU as preventing the UK from denying entry, refusing residence or deporting citizens from other EU Member States. This is plainly false. The EU Treaties explicitly provide that national authorities can limit the exercise of EU free movement rights on grounds of public policy, public security or public health. The main piece of EU legislation dealing with the rights of EU citizens and their family members to move and reside freely within the EU contains a whole chapter on ‘restrictions on the right of entry and the right of residence on grounds of public policy, public security or public health’ (see Chapter VI of Directive 2004/38).
In other words, and to cite the European Council, host Member States may ‘take the necessary restrictive measures to protect themselves against individuals whose personal conduct is likely to represent a genuine and serious threat to public policy or security. In determining whether the conduct of an individual poses a present threat to public policy or security, Member States may take into account past conduct of the individual concerned and the threat may not always need to be imminent. Even in the absence of a previous criminal conviction, Member States may act on preventative grounds, so long as they are specific to the individual concerned.’ (Decision of the Heads of State or Government, meeting within the European Council, concerning a new settlement for the UK within the EU – N.B. This statement merely summarises the current situation notwithstanding that it appears in the now defunct ‘new settlement’ following the outcome of the UK’s Brexit referendum)
Compliance with the rule of law and respect for human rights however mean that EU Member States must comply with a number of substantive and procedural safeguards provided by EU law. When it comes for instance to public policy or public security, national measures justified on these grounds must be proportionate and based exclusively on the personal conduct of the individual concerned, which must represent a genuine, present and sufficiently serious threat affecting one of the fundamental interests of society. EU law also provides for a number of procedural safeguards. For instance, any decision to deny entry or expel an EU citizen must be notified in writing and include precise and full information of the grounds on which the decision is based. In the absence of such safeguards, nothing would prevent national authorities from behaving arbitrarily and target certain individuals or groups for political or economic reasons.
While one may legitimately defend the view that the substantive and procedural safeguards provided by EU law are too protective of EU citizens and their family members, critics in most cases fail to make clear the extent to which the current safeguards should be lowered.
To argue however that the EU Court of Justice has prevented the UK from deporting 50 criminals from other EU Member States, as was argued by Vote Leave during the Brexit campaign, is simply false. The record on this issue has been set straight by Professor Steve Peers in this blog post.
(7) Does EU Law prevent Member States from imposing a ‘brake’ on EU immigration?
The short answer is yes as EU law currently stands. This certainly does not mean that EU Law provides for any ‘absolute’ right to move and reside freely within the territory of EU Member States. As described above in our answer to Question 1, EU free movement rights can only be exercised in accordance with the conditions and limitations laid down in the Treaties and the legal instruments adopted thereunder. And while the EU Treaties provide for general derogations as regards the principle of non-discrimination on grounds of nationality with respect to employment in the public service or activities connected with the exercise of official authority, any permanent quota or reliance on a ‘brake system’ regarding intra-EU migration would not be compatible with EU law. Any general ban on EU workers being treated equally with national workers as regards employment, remuneration and other conditions of work and employment would similarly breach EU Treaties.
The compatibility of a provisional, temporary ‘emergency brake’ on equal treatment with EU Treaties was however assumed by the European Council when it agreed last February to push for the adoption of a new ‘alert and safeguard mechanism’ in order to assuage the concerns of the UK government then led by David Cameron (this mechanism formed part of the ‘new settlement for the UK within the EU’ whose entry into force was however dependent on a decision by the UK to stay in the EU). This proposal was not for a brake on immigration, but on EU national workers’ access to in-work benefits.
This mechanism was supposed to work as follows:
(i) Any Member State facing an inflow of workers from other Member States of ‘an exceptional magnitude over an extended period of time’ could avail of the new alert and safeguard mechanism;
(ii) In order to be authorised by the Council to limit the access of newly arriving EU workers to non-contributory in-work benefits for a total period of up to four years, the Member State was expected to demonstrate the existence of ‘an exceptional situation’ affecting ‘essential aspects of its social security system’ or leading ‘to difficulties which are serious and liable to persist in its employment market or are putting an excessive pressure on the proper functioning of its public services’.
(iii) Assuming that such a demonstration had been offered, the Council could authorise the Member State concerned to restrict access to non-contributory in-work benefits to the extent necessary.
(iv) Finally, any authorisation would have a limited duration and the 4-year restriction on access to non-contributory in-work benefits could only apply to EU workers newly arriving during a maximum period of 7 years.
Contrasting legal views have been expressed regarding the compatibility of this new safeguard mechanism with the EU Treaties. In our opinion, this mechanism would have been vulnerable to legal action as its compatibility with EU Treaties is arguable in the absence of a Treaty amendment.
Notwithstanding this legal issue, we have always found it difficult to understand how the UK could rationally avail itself of such mechanism considering that it then enjoyed the highest rate of employment in UK history. Furthermore, evidence of any negative impact of EU work immigration to the UK on vulnerable workers and the sustainability of the UK welfare system was lacking (see this Financial Times article published on 22 February 2016). Furthermore, it has since emerged that the UK government did not in fact possess any hard evidence which would show the negative impact of EU migration to the UK and could justify the activity of the agreed ‘emergency break’ (see this article by the former deputy director of the policy unit’s in David Cameron’s government: ‘To be honest, we failed to find any evidence of communities under pressure that would satisfy the European Commission. At one point we even asked the help of Andrew Green at MigrationWatch, an organization that has been critical of migration. But all he could provide was an article in the Daily Telegraph about a hospital maternity ward in Corby. There was no hard evidence.’)
This EU safeguard mechanism is, in any event, no longer on the table. The UK could however seek to join the EFTA and remain part of the EEA post Brexit, which would offer the advantage of maximum access to the EU’s internal market with the additional option of availing of the special safeguard regarding free movement of people laid down in Article 112 of the EEA agreement:
1. If serious economic, societal or environmental difficulties of a sectorial or regional nature liable to persist are arising, a Contracting Party may unilaterally take appropriate measures under the conditions and procedures laid down in Article 113.
2. Such safeguard measures shall be restricted with regard to their scope and duration to what is strictly necessary in order to remedy the situation. Priority shall be given to such measures as will least disturb the functioning of this Agreement.
3. The safeguard measures shall apply with regard to all Contracting Parties.
Assuming that the UK aims and is able to remain part of the EEA after its eventual exit from the EU, triggering this provision with the view of limiting EEA labour immigration to the UK would not however be cost-free. Indeed, and to the best of our knowledge, none of the contracting parties to the EEA has ever done so with respect to labour immigration if only because another provision of the same agreement entitles parties negatively affected by safeguard measures to take ‘proportionate rebalancing measures’ (the special regime granted to Liechtenstein with regard to nationals of EU Member States and EFTA States, which will be briefly mentioned below, is not directly based on Article 112 EEA).
In other words, should the UK as an EEA member trigger this provision in order to impose quantitative limitations or restrictive rules regarding entry, residence and employment of EEA nationals in the UK, it could then be subject to what would be essentially retaliatory measures focusing for instance on UK banks’ passporting rights (i.e. rights derived from EU law which enable banks based in the UK to offer financial services to the rest of the EEA without having to follow the regulations of the countries where they are offered). Furthermore, the UK would have to accept that any dispute in this context may be referred to an arbitration panel (see Protocol 33 on arbitration procedures of the EEA Agreement).
In addition to this EEA general safeguard mechanism, the EU and EFTA countries were able to agree a rather unique arrangement regarding Liechtenstein with respect to free movement of people, which was initially supposed to be provisional (see Protocol 15 on transitional periods on the free movement of persons) but ended up becoming de facto permanent (for more details see this 2015 Communication from the European Commission). In a nutshell, due to its specific nature as a microstate with a population of 36,925 with an already unusually high percentage of non-national residents and employees, Liechtenstein was permitted to introduce quantitative limitations (aka a quota system) to control the number of EEA citizens wishing to take up residence via a system limiting the number of residence and short-term permits a year. In 2015, the Commission concluded that there is ‘no need to make any changes to the current rules’ until another review of this specific arrangement takes place in 2019.
It has been suggested that a Liechtenstein-inspired solution may be the way forward as far as the UK is concerned (see the proposal by Vicky Ford, a Conservative MEP as reported in The Guardian). This is highly doubtful if only because the quota system agreed for Liechtenstein was explicitly justified on the basis of this country’s ‘specific geographic situation’ and a total number of employees which is almost equal to the number of residents, 52 % of whom commute from neighbouring countries, a situation in no way comparable to the UK’s (see Question 10 below for an overview of other possible options for the UK to consider).
(8) Does EU Law only guarantee a right of permanent residence after five years?
There has been a fair amount of confusion on the EU right of permanent residence in the UK. To put it briefly, it is regularly argued that such a right would only arise after five years of residence but this is not entirely accurate.
This point is not easy to explain briefly. To begin with, it is correct to state that EU Law provides for a right of permanent residence for EU citizens and their family members after five years of continuous legal residence in the host EU country. This right directly derives from EU law and is explicitly mentioned in Directive 2004/38.
What is usually misunderstood however is that EU citizens and their family members (irrespective of nationality) can acquire a right of permanent residence, in the sense of continued and uncontested – but conditional - residence as soon as they move and reside in a different Member State. In other words, provided that an EU citizen maintains his/her status as worker or self-employed person, or that he/she have sufficient resources for himself/herself (and eventual family members) as well as a comprehensive sickness insurance cover, EU law guarantees a right of continued residence in the host Member State.
The added value of the new right of permanent residence after five years of residence laid down in Directive 2004/38 lies in the fact that once it is obtained, this right is no longer subject to the conditions previously mentioned.
To summarise, EU law guarantees a conditional right of permanent residence before five years and a nearly unconditional one after five years (it may be lost through continuous absence from the state for over 2 years; and there are powers to remove criminals from the country as discussed in the answer to question (6) above).
To claim that only those having resided in the UK for more than five years are entitled to claim a right of permanent residence misrepresents EU law as it stands. This is an important point considering that it is regularly reported in the press that the UK government may only seek to allow EU citizens with a right of permanent residence to stay in the UK by the time the UK leaves the EU, while those with allegedly no permanent residence right by then would be offered an ‘amnesty’ (see this article from The Telegraph published on 7 October 2016).
This reflects a misunderstanding of EU law. As noted above, and just to give a single example, any EU worker residing in the EU is entitled to reside permanently in the UK from day 1 of his/her residence in the UK as long as he/she maintains his/her status of EU worker. It might be clearer to propose that rights attach to ‘Article 16’ (of the EU citizens’ Directive) or ‘unconditional’ permanent residence, rather than just ‘permanent residence’.
In any case, if the yardstick is those with an unconditional right to reside thanks to the Directive 2004/38 concept of permanent residence, then the threshold is not always five years. There are exceptions for instance for those who retire in the host member state (the threshold is 3 years of residence, if they have been working in the host state for at least the past 12 months), or if they have to stop working as a result of permanently incapacitating illness (the threshold is two years), or if they become incapacitated as a result of an accident at work or occupational disease (and are entitled to a relevant benefit from the host state), in which case there is no duration of residence requirement. It is therefore not only EU nationals who have resided for five years who have a ‘clear cut’ right to reside.
One should also note in passing that to speak of amnesty in this context is to deeply misunderstand the legal meaning of this concept, which either refers to a general official pardon for people, convicted of political or criminal offences or an undertaking by public authorities to take no action against specified offences during a fixed period. To speak of ‘amnesty’ to address the situation of EU nationals who have exercised their free movement rights to move to the UK in full compliance with the conditions laid down in EU Law, creates an unfortunate and misleading depiction of lawfully resident and working EU nationals.
(9) Would EU Law or International Law protect ‘acquired’ free movement rights following a British exit from the EU?
First of all, it is worth noting that we are talking about protection of rights in the event of a ‘repeal’ of those rights – i.e., if legislation in the UK is amended to remove free movement rights following the formal withdrawal of the UK from the EU. It is worth pointing out in this respect that counsel for the government in the litigation on whether triggering Article 50 requires parliamentary consent, has argued that the rights of EU nationals do not simply fall away on exiting the EU. Those rights are part of domestic law, implementing EU law. It is not impossible that the UK could leave the EU and leave all the free movement rights in tact. However, we face a question of what happens if it does not.
Prior to the referendum, there have been repeated suggestions that British citizens in the EU and EU nationals in the UK would be able to rely upon the Vienna Convention’s protection of ‘acquired rights’ to enable them to stay and to protect their status there.
Legally speaking, this is problematic and highly unlikely. The Convention in question – the Vienna Convention on the law of the Treaties – contains a provision, Article 70(b), which states that the termination of a treaty:
“Does not affect any right, obligation or legal situation of the parties created through the execution of the treaty prior to its termination.”
However, this provision refers to rights of ‘the parties’. The parties to the EU Treaties are Member States, not individuals. It was therefore wrong for Gisela Stuart or Boris Johnson to imply that Article 70 of the Vienna Convention directly addresses individual free movement rights acquired Pre-Brexit. For instance, the International Law Commission had for instance made clear that this provision is simply not concerned with the acquired or vested rights of individuals. This is why Jean-Claude Piris (former General Director of the Legal Service of the Council of the EU) described the Vienna Convention claim as a ‘new legal theory’ with no ‘legal support in the Treaties’, which must be disregarded as it ‘would lead to absurd consequences’.
To summarise, while international law does recognise the notion of ‘acquired rights’, there is, as observed by Professor Vaughan Lowe QC a ‘general agreement that the category of ‘acquired rights’ does not extend beyond property rights and certain contractual rights’, which means that the EU ‘rights to live, work, receive medical care and retire in an EU Member State other than one’s own (or for companies, the right of establishment) would not be included within that category.’
In any case, claiming a continued right of residence is a claim for a present and future right, not an acquired one. It would be strange were it possible to claim continued reliance upon EU Treaties which have ceased to apply, given that Article 70 (a) of the Convention states that the termination of the Treaty “Releases the parties from any obligation further to perform the treaty”. The sorts of past State rights that might be protected by the Convention could be that the EU could not demand that the UK pay back Structural Funds paid to correspond with past periods during which the UK was a member of the EU, where those funds were paid in good legal order.
And it is worth pointing out that the Convention is a UN instrument, and there is little ground for supposing that it would be readily actionable to claimants facing negative administrative decisions, since it does not create direct rights. Nor is it able to trounce both EU and national laws in other Member States. If it were able to do so, that would raise significant questions of supremacy, sovereignty and transparency.
During the course of the campaigns, Matthew Elliot of Vote Leave also suggested that the EU Charter of Fundamental Rights would protect British nationals’ rights to stay in the EU. However, (a) the Charter is silent on the matter of rights of ex-EU nationals; (b) it does not create any additional rights not already in existence within EU law; and (c) The Charter is an EU instrument. If the UK exit agreement includes provisions on UK nationals in the EU (and EU nationals in the UK), then those provisions must be interpreted in accordance with the Charter – e.g. in accordance with a right to family life. Outside of the agreement, these individuals may anyway fall within the scope of EU law, e.g. through having married a national of a Member State who has exercised their right to free movement, or being the primary carer of a host State national. In any event, if the UK was no longer bound by the Charter as a non-member, then UK nationals in the EU seeking to rely upon it would have to show that their situation fell within the scope of EU law (such as EU legislation on migration from non-EU countries) for the Charter to apply, and then show that a fundamental right covered by the Charter was engaged.
The key point here is that international law relating to ‘acquired rights’ does not offer any solid basis for a claim to retain EU rights of residence post-Brexit. And this is why the UK government, without mentioning the Vienna Convention, has decided that the criteria which would ‘enable EU citizens to remain in the United Kingdom following exit from the European Union will depend on the outcome of the negotiations and the scope of any reciprocal agreements concerning British citizens who live in other member states.’ (Lord Keen of Elie, 29 June 2016)
While there are no clear legal guarantees for either the residence rights of UK nationals in the EU or of EU nationals in the UK, there are nevertheless strong legal arguments in favour of offering protections to EU nationals in the UK. The ‘bargaining chip’ stance is dehumanising and ethically problematic; it puts the lives of nationals from other EU countries living in the UK ‘on hold’ for an indefinite period of time (with very practical consequences such as increased difficulties to get loans or mortgages, not to mention the anxiety it has created). This is why, for instance, Michael Howard, the former Conservative leader, called on Mrs May to ‘lead by example’ end the ‘dreadful uncertainty’ facing EU migrants living in the UK.
That being said, the suggestion implicitly made by David Davis that EU nationals could be deported retrospectively (i.e. even if they arrived before the UK leaves the EU) may, in circumstances where their right to family life was at stake, be open to legal challenge on the basis of the UK’s Human Rights Act. See e.g. Matthew White’s post on ‘When can EU citizens be expelled from the UK after Brexit?’ available here and Camino Mortera-Martinez & John Springford’s CER insight piece ‘Britain will struggle to make EU migrants ‘go home’’ available here. Equally, UK citizens in the remaining EU could invoke the right to family and private life in the European Convention on Human Rights, as well as rights in national constitutions, even if they were not covered by EU immigration law or transitional arrangements. It should be noted however that winning an Article 8 ECHR claim in the immigration courts it is not straightforward.
(10) What are the main options available to the UK government?
It is not yet entirely clear whether the UK government wishes to revise or completely reject the application of current EU rules in the UK. A complete rejection would not be compatible with an extensive access to the EU’s single market à la norvégienne (see below for more details on this model).
There is however a large spectrum of different free movement options that are, in theory, possible for the UK to consider but the default (political) principle may be summarised as follows: The more extensive the UK restrictions on the free movement rights of EU citizens and family members, the less extensive the UK’s access to the EU’s single market is likely to be. This was the point made by the German Chancellor, Angela Merkel, when she stated that full access to the European single market depended on whether on the acceptance of the EU’s four freedoms, including the freedom of movement of people: “If Britain says no, it can’t get full access to the European single market”.
Assuming the UK wishes to retain as much access to the European single market as possible, it would then have to seek membership of the EEA – the so-called Norway model. This involves not being an EU member, but applying much of EU law, including provisions on free movement, equal treatment on the ground of nationality, and social security coordination. Some have suggested, slightly further along the spectrum, a Norway-minus model, that adapts the EEA model, to allow for more restricted movement of workers, perhaps in return for ‘a bit less single market’ access. Alternatively, the UK could instead join the EEA and seek to trigger its safeguard mechanism as noted above in our answer to question 7.
The Swiss bilateral agreement model is another option that involves free movement but permits greater restrictions on equal treatment with regard to welfare benefits. The model is complicated, relying on a ‘patchwork’ of agreements that need regular updating. The relationship is not entirely stable – calls to renegotiate the free movement provisions have been rejected by the EU; a Swiss quota initiative has soured the EU-Swiss relationship, in the light of which the Swiss government appears to be considering a second referendum which could ask voters to decide between the unilateral imposition of curbs on EU migrants and maintenance of Switzerland’s current access to the EU’s single market (another alternative plan is to sidestep ‘quota in favour of giving current Swiss residents labor-market precedence’ but this idea is unlikely to be agreed by the EU).
The Turkish system is much closer to the other end of the spectrum, since there is no free movement as such, but there are some special arrangements, such as the accrual of increasingly ‘equal’ employment rights. Turkish workers legally employed in the EU are entitled to the same working conditions as EU nationals, and also have a sliding scale of rights: after one year of legal employment they are entitled to have their work permit renewed if a job is available; after three years of legal employment they are entitled to switch employers and respond to other job offers within the same occupation; after four years of legal employment they have free access to any paid employment in that EU country. Borrowing from this approach could involve heavy restrictions – such as requirements for residence and work permits – but allowing EU workers, once they had been accepted into the UK, to accrue residence, employment and social rights during their time here.
Or it is possible to envisage ad hoc models at different points along the spectrum. One such is a Continental Partnership, involving a reciprocal quota system – the UK imposing a quota system on EU entrants, and the EU imposing a quota system on UK entrants. This is problematic, since it is not clear how this could be made workable, partly in light of the substantial cross border populations already in situ, and the migration flows that stem therefrom, and also because of the fluidity with which people would likely shift status (assuming the system did not apply to visitors). The idea of a quota imposed on the EU as a whole would be tricky to administrate fairly – how could/should a system deal with a quota being ‘used up’ by one or more states disproportionately? And should UK nationals be denied the possibility to move to Poland because the EU quota has already been met by UK nationals in Germany? The EU Treaties also state that quotas on non-EU citizens coming to work are decided by Member States individually (Article 79 TFEU).
A variation on this idea is ‘preferential movement’ – allowing free movement for immigrants above a set salary or skill level, and setting quotas below it – primarily because that accords with ‘UK public attitudes to immigration’, rather than with economic evidence. It would likely be administratively and legally complex to set the categories, then maintain two parallel immigration systems for the same nationality cohort and monitor continued status for each individual within the allocated category. Furthermore, it would seem irresponsible for public policy to be driven by popular ‘feelings’ about the impact of immigration rather than by evidence of the impact of immigration, especially in a context where the public may have been misled by media misrepresentation (see e.g. this article in The Economist on the EU myths propagated by the British media).
Another idea floated recently was that of a ‘job-first’ model. This is also not without problems. It is not clear that it would reduce immigration (which is the apparent purpose), but would simply place the recruitment process one step back. This could give more power to gang-masters, who already have the machinery for cross-border recruitment in place, and are often linked to employers who provide tied accommodation, and transport their workers to and from the place of work. This could lead to EU nationals being imported in groups, and living quite controlled lives, segregated from much of the UK population.
The other scheme apparently debated by the UK cabinet, which one may label ‘targeted work permit system’, is similarly problematical. As reported by the Financial Times, this system would ‘most likely allocate a set number of visas by sector’ for highly skilled workers in receipt of a job offer. By contrast, unskilled workers could only gain entry into the UK on the basis of a temporary workers scheme, which would grant EU unskilled workers access to the UK for a fixed period of time with no accrued residency rights. Such a ‘targeted work permit system’ however assumes that one can clearly define what is an ‘unskilled worker’ as opposed to an ‘highly skilled’ one. It would be incompatible with access to the European single market in a similar way to Norway or Switzerland and would be incredibly administratively complex.
Others have suggested the devolution of immigration policy as far as EU nationals are concerned. This would mean for instance giving Scotland, Northern Ireland but also London and Gibraltar the right to continue to apply current free movement rules to EU nationals and family members (EEA membership for Scotland would make this a compulsory feature as rightly noted in this Financial Times article by Martin Sandbu) or alternatively, to apply a different but more liberal immigration system on a regional basis (see e.g. the ‘regional visa system’ proposal made by the City of London which is available here). However, it is difficult to see how such systems could be made tenable, without introducing intra-national migration controls.
In our opinion, the UK government should not seek to fundamentally undermine the current EU’s labour market-based system of intra-EU migration control, where evidence suggests that the system works well for the UK – a system which both the Japanese government and the US Chamber of Commerce have described as key for attracting and maintaining foreign investment in the UK. Instead of devising cumbersome and costly alternative schemes, it may be more appropriate to instead aim to revisit the question of restrictions already permitted within EU law, while also questioning whether the ‘mischief’ that these proposals are intended to address actually exists.
Barnard & Peers: chapter 13, chapter 27
Photo credit: LSE blog
dr. Agne Limante
MA candidate in EU Law, King’s College London; PhD in EU law, Vilnius University; Post-Graduate Diploma in EU Law, King’s College London
When in November 1993 the CJEU presented its ruling in Keck case[i], seeking, inter alia, ‘to clarify the law’, academic society reacted to it with a plethora of studies on how to ‘clarify the clarification’[ii]. In that ruling the Court held that French rules prohibiting reselling at loss were not caught by Article 34 TFEU since they related merely to indistinctly applicable ‘selling arrangements’ having no adequate effect on intra-EU trade.
It was clear that the judgement signalled contradiction in the scope of application of EU rules on free movement of goods. As noted by prof. Weatherill, Keck was doubtless intended to empower national courts to dismiss far-fetched attempts to deploy internal market law which was clogging up the EU judicial system with the minutiae of purely local affairs[iii]. However, the borders of such limitation and its content remain blurred until today.
The analysis below will re-think Keck from the position of market access test. It firstly shortly examine the nature of ‘selling arrangements’ introduced by Keck, and later will turn to the discussion on the market access test as an alternative approach to Keck.
Restrictive measures falling under Keck
According to the settled case-law, all trading rules enacted by the Member States which are capable of hindering, directly or indirectly, actually or potentially, trade within the EU are to be considered as measures having an effect equivalent to quantitative restrictions and are, on that basis, prohibited by Article 34 TFEU[iv]. However, according to Keck, the application to products from other Member States of national provisions restricting or prohibiting certain selling arrangements, within the Member State of importation, is not such as to hinder trade between Member States so long as, first, those provisions apply to all relevant traders operating within the national territory and, secondly, they affect in the same manner, in law and in fact, the marketing of domestic products and those from other Member States.
Following the above, while retaining the prohibition of hindrance to market access set in Dassonvilleand the general non-discrimination requirement, Keck made a distinction between:
(i) ‘rules that lay down requirements to be met by goods’, related to inherent characteristics of products, and per se are considered to fall under Article 34 TFEU;
(ii) ‘selling arrangements’, regulating questions extrinsic to goods. They are covered by Article 34 TFEU only if they discriminate against imports;
(iii) ‘residual rules’, as a separate group named in subsequent case-law and doctrine. They fall under Dassonville as affecting intra-EU trade.
Identifying selling arrangements, however,appeared to be a puzzle. Although they seem to cover restrictions on when[v], where and by whom[vi] the goods may be sold, as well as advertising restrictions[vii] and price controls[viii], these general elements hardly allow predicting outcomes in individual cases.
It should also be noted that during last decades the Court increasingly adopted a broad interpretation of indirect discrimination bringing the rules on marketing within the scope of Article 34 TFEU. Indeed, as stated by prof. Spaventa, aside from Sunday-trading type of rules, during the period of 1997-2008, there were only two cases – Burmanjer and A-Punkt – where selling arrangements were found to be non-discriminatory and thus fell outside the scope of Article 34 TFEU[ix].
Market access test as an alternative to Keck
Since Keck received a lot of criticism, there were considerable attemps to reformulate it and to suggest alternatives. The majority suggested market access test to be the panacea.
As Oliver and Enchelmaier[x] note the most authoritative assault ever mounted against the reasoning in Keck was that of AG Jacobs in Leclerc-Siplec[xi]. He considered that it was inappropriate to make such a rigid distinction between “rules that lay down requirements to be met by goods” and ‘selling arrangements’, and that the test of equality is not in line with the objectives of the Treaty[xii], namely, the establishment of a single market. At the same time AG Jacobs suggested a test aimed at checking whether a measure exerts a substantial restriction on the market access[xiii]. Since Leclerc-Siplec related to partial ban on advertisement (TV advertising), he came to a conclusion that this did not amount to substantial restriction of the market access. However, he also offered an example of direct TV marketing: here prohibition of TV advertising would mean substantial restriction. Comparably, in his opinion in Alpine Investments[xiv] dealing with restriction on ‘cold calling’ potential clients offering financial services, AG Jacobs argued that a question whether a national rule restricts freedom to provide services “should be determined by reference to a functional criterion, that is to say, whether it substantially impedes the ability <…> to provide services”.
Prof. Weatherill[xv] chose another direction stating that prohibition to resale at loss in Keck escaped Article 34 TFEU not because these were rules affecting selling arrangements applied equally in law and in fact, but because they were measures applying equally in law and in fact and exercising no direct impediment to access to market of a Member State. There was no obstruction to the realization of economies of scale and wider consumer choice. Going further, prof. Weatherill also sugests refining Keck test along the following wording: measures introduced by a Member State which apply equally in law and in fact to all goods or services without reference to origin and which impose no direct or substantial hidrance to the access of imported goods or services to the market of that Member State escape the scope of application of Articles 34 and 66 TFEU. Following such suggestion, complete bans on sale of goods or services (as in Schindler[xvi]), even if equally applied, would still have to be justified for having direct and substantial hidrance to market access.
AG Maduro in Alfa Vita[xvii] also stressed market access. He brought to the attention three criteria to be used when deciding on application of Article 34 TFEU. First, discrimination based on nationality must be prohibited. Second, the imposition of supplementary costs on cross-border activity has to be justiﬁed. Lastly, the measure will be a hidrance to market access if it impedes to a greater extent access to the market and putting into circulation of products from other Member States than trade on national market.
However, other authors do not see much use in notions such as ‘substantial barrier’ or in introducing other types of treshold. They claim that this would introduce an unwanted de minimis test, while ‘barrier to market access’ criterion is inherently nebulous[xviii]. Even if it is possible to distinguish restrictive rules that have minor effects on intra-EU trade, such as ‘modest effect on sales’,‘purely hypothetical and totally uncertain and indirect effect on market access’[xix], a large grey area is left between ‘minor’ and ‘substantial’.
Use restrictions, Keck and market access
Following Moped trailers[xx] and Mickelsson & Roos[xxi], the contours of Keck seem even more blurred and, again, subject to various interpretations. Both cases were related to the limits of market access and restriction of use – prohibition of towing the motor trailers and prohibition on use of personal watercrafts respectively.
These cases reveal the complexity of the question. AG Bot in Moped trailers case argued that the extension of Keck to measures regulating the use of goods would run against the aims of the internal market. AG Kokot, on the opposite, in Mickelsson & Roos suggested to extend Keck criteria to the use arrangements due to comparable characteristics.
In Moped trailers CJEU re-defined the notion of barriers to intra-EU trade underlining the market access. It stated that “any other measure which hinders access of products originating in other Member States to the market of a Member State is also covered by the [measures having equivalent effect] concept“. The Court reaffirmed this market access formula in Mickelsson & Roos.
Some authors guessed that with this case law the Keck selling arrangements doctrine might have been consigned to history books: though the Court did not openly overrule Keck, the market access formula might suggest, in fact if not in law, the end of the Keck dichotomy[xxii]. The Court made clear that any measure which may impede access to the market falls under Article 34 TFEU. Although distinguishing product-related rules remains necessary, delimitation of selling arrangements looses significance.
It is, however, a bit suprising that the suggested test established no type of threshold or qualifiers to be met by national measure. On the other hand, as noted by prof. Snell, the CJEU seems to focus on the significance of the impact of the measure with all the uncertainties this approach entails, though at the same time refusing to clearly state that rules with insignificant effect fall outside the Treaty[xxiii].
Looking from the other side, confronting Keck and market access test, we might be looking for opposition where it does not exist. Keck is about market access. As stated by Wenneras & Boe[xxiv], the CJEU in Moped trailers and Mickelsson & Roos has just consolidated and clarified what was implicit in Keck, namely that Article 34 TFEU prohibits measures that discriminatorily, in law or in fact, restrict market acess for imported products or which prevent/hinder market access.
Nevertheless, it is often hard to identify, which of the elements plays the decisive role in CJEU judgements – market access, discrimination and protectionism, economic freedom. One can have an impression that those tests remain elusive and are often used in an intuitive way, after firstly looking whether discrimination could be established, having in mind the possible justification in certain case and combining this intuition with the general feeling of what is reasonable and logical.
[i] Joined Cases C-267/91 and C-268/91 Keck and Mithouard  ECR I‑6097.
[ii] Weatherill. After Keck: Some thoughts on how to clarify the clarification. (1996) 33 CMLRev, p. 885-906.
[iv] Case 8/74 Dassonville  ECR 837.
[v] Case C-401/92 and C-402/92 Tankstation ’t Heukskeand Boermans  ECR I-2199, Case C-69/93 and C-258/93 Punto Casa and PPV  ECR I‑2355.
[vi] Case C-391/92 Commission v Greece  ECR I‑1621, Case C-387/93 Banchero  ECR I-1085. It is worth noting, that restrictions related to where the product might be sold are captured by selling arrangements if they relate to restrictions on physical location of selling place (pharmacy-shop-district). Selling by internet is given a different treatment. In DocMorris (C-322/01,  ECR I-14887), case concerning prohibition of selling medicines by virtual pharmacies, the CJEU hold the restriction to be de facto discrimination against imports. Internet provided a “more significant way” to market access for sellers established in other Member States.
[vii] Case C-292/92 Hünermund and Others  ECR I-6787; Case C-412/93 Leclerc-Siplec  ECR I-179; Cases C-34/95 to C-36/95 De Agostini  ECR I-3843, Case C-405/98 Gourmet  ECR I-1795. However, often advertising restrictions might have discriminatory aspect. As AG Jacobs insisted in Leclerc-Siplec, “measures that prohibit or severely restrict advertising tend inevitably to protect domestic manufacturers and to disadvantage manufacturers located in other Member States“.
[viii] Case C-63/94 Belgapom  ECR I-2467.
[ix] Spaventa. Leaving Keck behind? The free movement of goods after the rulings in Commission v. Italy and. Mickelsson and Roos. (2009) 34 ELRev. 914-932.
[x] Oliver & Enchelmaier. Free Movement of Goods: Recent Developments in the Case Law. (2007) 44 CMLRev. 649–704.
[xi] Case C-412/93 Leclerc-Siplec  ECR I-179.
[xii] In AG Jacobs words: “If an obstacle to inter-State trade exists, it cannot cease to exist simply because an identical obstacle affects domestic trade”.
[xiii] It is interesting to note, that AG Jacobs in general tends to analyse measures in the light of the scope of their effect. Criticizing locus standi in annulment actions (UPA case), he also suggested a test of “substantial negative effect” for interpretation of individual concern.
[xiv] Case C-384/93 Alpine Investments  ECR I-1141.
[xv] Weatherill. After Keck: Some thoughts on how to clarify the clarification. (1996) 33 CMLRev, p. 885-906.
[xvi] Case C-275/92 Schindler  ECR I‑1039.
[xvii] Joined Cases C-158/04 and C-159/04 Alfa Vita  ECR I‑8135.
[xviii] Oliver & Enchelmaier. Free Movement of Goods: Recent Developments in the Case Law. (2007) 44 CMLRev. 649–704.
[xix] Joined Cases C-418/93, etc Semeraro Casa Uno  ECR I‑2975.
[xx] Case C‑110/05 Commission v Italy  ECR I‑519.
[xxi] Case C‑142/05 Mickelsson and Roos  ECR I‑4273.
[xxii] Spaventa. Leaving Keck behind? The free movement of goods after the rulings in Commission v. Italy and. Mickelsson and Roos. (2009) 34 ELRev. 914-932.
[xxiii] Snell, The Notion of Market Access: A Concept or a Slogan? (2010) 47 CMLRev, p. 437-472.
[xxiv] Wenneras & Boe Moen. Selling arrangements, Keeping Keck. (2010) 35 ELRev. 387 -400.