EU Referendum

Brexit: free movement and the Single Market


This is an updated and extended post on the Liechtenstein/EEA solution on "free movement of persons", bringing new material and sources, making for as complete a record as possible.

For the forthcoming exit negotiations between the United Kingdom, it is generally regarded as an absolute that the UK's continued participation in the Single Market is dependent on acceptance of all four freedoms written into the EU treaties, including the free movement of persons.

That much was made clear by European Council President Donald Tusk at the informal meeting of the 27 EU Member States (minus the UK) on 29 June 2016. He added: "There will be no single market à la carte", thereby adding his name to a long list of EU officials and Member State politicians who have indicated that changes to freedom of movement are "non-negotiable".

This includes Angela Merkel who recently said during a speech at the annual diplomatic corps reception in Meseberg, north of Berlin, that: "… whoever would like to have free access to the European internal market will also have to accept all basic freedoms in return, including the free movement of people".

However, this lack of flexibility may have more to do with political posturing than reality. The European Commission, by its own account, has "always stressed that free movement was a qualified right and not an unconditional one".

This was in the wake of the Dano case in the European Court of Justice, but referred back to an earlier case where the Court had declared that the treaties and secondary legislation had "qualified and limited" freedom of movement.

The point that emerges is that there is nothing absolute, in principle, about freedom of movement. Therefore, there is no legal bar to variations being negotiated, given the political will. Furthermore, it is the case that the Union has been prepared both to negotiate and compromise on this issue.

Specifically, these negotiations lie within the domain of the EEA Agreement, related to (but not necessarily entirely reliant on) the "safeguard measures" set out in Articles 112 and 113.

It is essential, therefore, that we look fully and openly at the possibilities of the UK seeking a compromise on Single Market participation which will permit national limitations or restrictions on freedom of movement (i.e., immigration) of citizens from EU Member States.

This could be done within the framework of the EEA Agreement, the best example of which is the so-called "Liechtenstein solution". In this post, we look at this solution, bringing it right up to date, and then the wider legal and political issues, and draw conclusions for the record, to ensure that the issues are openly are fully rehearsed.

The Liechtenstein solution: "sectoral adaptation"

Prior to the principality of Liechtenstein joining the EEA on 1 May 1995, the EEA Council – one of the formal structures set up under the agreement – on 10 March 1995 looked at its vulnerability to excessive migration.

It concluded that this microstate could easily be swamped by immigrants if unrestricted free movement of workers was permitted. A territory with a population of 37,000 spread over an area of 61 square miles – less than half the area of the Isle of Wight – would not be able to absorb unlimited numbers.

The Council recognised that Liechtenstein had "a very small inhabitable area of rural character with an unusually high percentage of non-national residents and employees. Moreover, it acknowledged the vital interest of Liechtenstein to maintain its own national identity". It thus concluded that the situation "might justify the taking of safeguard measures by Liechtenstein as provided for in Article 112 of the EEA Agreement".

With that, it asked the Contracting Parties to "endeavour to find a solution which allowed Liechtenstein to avoid having recourse to safeguard measures". However, no long-term solution was found so a temporary expedient was arranged by way of transitional arrangements which allowed the country to impose "quantitative limitations" on immigration until 1 January 1998. These were incorporated into Protocol 15, appended to the Agreement.

Towards the end of 1997, just before the end of the transitional period, there had been no further measures proposed so Liechtenstein unilaterally invoked the Article 112 safeguard measures, thereby continuing to keep the existing immigrations restrictions in place when the transitional period ended.

There were further attempts to resolve the situation in 1998, which were unsuccessful. Then, on 17 December 1999 after a further review, the EEA Joint Committee decided that the "specific geographical situation of Liechtenstein" still justified "the maintenance of certain conditions on the right of taking up residence in that country".

In order to resolve the situation, though, it came up with the proposal for a longer-term solution, allowing Liechtenstein to introduce a quota system controlling the number of workers allowed to enter the country.

This decision was given formal status by an amendment to Annex VIII of the EEA Agreement, setting out what were called "sectoral adaptations", cross-referred to Annex V on the free movement of workers.

The decision provided for a new transitional period until 31 December 2006, and introduced a formal amendment to the EEA Agreement, which allowed for the new measures to apply subject to a review "every five years, for the first time before May 2009". After reviews in 2009 and in 2015, it was concluded that there was no need to make any change to the current rules. The Sectoral Adaptations could remain unchanged.

Under the current arrangement, Liechtenstein issues a limited number (less than 100) of residence permits for economically active persons and a very much smaller number for economically non-active persons.

Half of the totally available permits are decided by lottery, held twice a year. The numbers involved are, of course, small beer, but Liechtenstein is a tiny country. What matters is that a precedent has been set within the framework of the EEA Agreement for suspending freedom of movement in respect of a single country, and replacing with a quota system for what amounts to an indefinite period.

It matters not that Liechtenstein is a micro-state. It is a fully-fledged contracting party within the terms of the EEA Agreement. What applies to one legally can apply to any or all.

Whatever the EU might declare in terms of freedom of movement being "non-negotiable" for EU Member States, therefore, it is undeniable that it is negotiable within the framework of the EEA Agreement, as it applies to Efta states. Therefore, it would appear that the scope exists to agree modifications to the principle of unrestricted freedom of movement, as did Liechtenstein, or unilaterally invoke Article 112 to achieve the same effect.

Safeguard measures

One important point to emphasise here is that while the EEA "safeguard measures" are a mechanism by which changes to freedom of movement could be secured, they are NOT relied upon by Liechtenstein for its current settlement.

With that caveat, it is worth looking briefly at the nature and application of safeguard measures, in general and specifically in relation to the EEA Agreement.

The point about safeguard measures generally is that, far from being rare and exceptional, they are commonly found in trade agreements. They can be found in the draft agreement with Australia and New Zealand, in the trade agreement with Moldavia and, in 1993, when Hungary signed up to an Association Agreement with the EU, Council Regulation No 3491/93 of 13 December 1993 detailed the procedures for applying safeguard measures.

As to the current safeguard measures in the EEA agreement, these are remarkably similar to the arrangements in Council Regulation (EEC) No 2840/72 of 19 December 1972, setting out the Agreement between the European Economic Community and the Swiss Confederation. Quite possibly, the EEA text is based on these provisions.

The EEA safeguard measures themselves can be triggered "If serious economic, societal or environmental difficulties of a sectorial (sic) or regional nature liable to persist are arising". And although such measures have to be "restricted with regard to their scope and duration to what is strictly necessary in order to remedy the situation", there is no specific time limitation.

This contrasts with the only safeguard measures written into Chapter 4 of the Treaty of the European Union (Article 66, TEU) on Capital and payments, which states:
Where, in exceptional circumstances, movements of capital to or from third countries cause, or threaten to cause, serious difficulties for the operation of economic and monetary union, the Council, on a proposal from the Commission and after consulting the European Central Bank, may take safeguard measures with regard to third countries for a period not exceeding six months if such measures are strictly necessary.
The comparison is highly instructive. While Article 66 TEU is self-evidently an emergency measure for highly specific situations, the safeguard measures in the EEA Agreement are most emphatically not an "emergency" provision.

The rules for its use, set out in Article 113, state that it cannot normally be used without first giving at least one month's notice. Only in "exceptional circumstances" can immediate action be taken, and then only to the extent "strictly necessary to remedy the situation". Logically, any provision which has within it an "emergency clause", for use only in exceptional circumstances, cannot in itself be an emergency measure.

Furthermore, while Article 66 is restricted to the highly specific issue of movement of capital which might "cause, or threaten to cause, serious difficulties for the operation of economic and monetary union", Article 112 of the EEA Agreement is much more broadly defined.

To trigger the Article, the Contracting Party can draw on three areas, defined as: "serious economic, societal or environmental difficulties". These can be of a sectorial (sic) or regional nature and the only limiting qualification is that they must be "liable to persist" – the very antithesis of a short-term crisis situation. On this basis, it can be said that "safeguard measures" can be emergency provisions, they are not necessarily so. It depends on the context.

Broader applications of the EEA Safeguard Measures

It has been asserted by Monnet Professor Dougan that the EU has been content to allow "adaptations" to freedom of movement to apply to Liechtenstein only because of its "nature, and the size and the territorial aspects".

However, as is evident from the 1994 Protocol adjusting the EEA Agreement, the original opt-out from freedom of movement provision, implemented under Protocol 15 to the EEA Agreement, applied to both Switzerland and Liechtenstein.

Nor were the safeguard measures confined to freedom of movement. In 1992, when the EEA Agreement was signed, the Final Act records that safeguard measures were invoked by no less than four of the (then) seven Efta members. Austria, Iceland and Switzerland cited the need to protect real estate, capital and labour markets.

The Government of Liechtenstein invoked Article 112 in respect of capital inflows, concerns about access of the resident population to real estate, and "an extraordinary increase in the number of nationals from the EC Member States or the other Efta States, or in the total number of jobs in the economy, both in comparison with the number of the resident population".

It is a matter of record that, after a referendum, the Swiss government was unable to ratify the EEA Agreement and its name was removed from Protocol 15. Had Switzerland not failed to ratify, the likelihood is that both Switzerland and Liechtenstein would currently enjoy exclusion from freedom of movement. Certainly, unilateral safeguard measures are currently being sought by the Swiss Federal Government as a resolution to the 2014 referendum on limiting immigration.

Such a solution has recently been looked-upon favourably by Martin Schultz, President of the European Parliament. He said that the idea of a so-called "safeguard clause", which has been thrown around among members of the Swiss government and parliament as a possible solution, seems promising at first glance. Such a clause, he said, "would introduce quotas after a certain immigration threshold is achieved in specific regions and industries".

As regards Iceland, having recorded its intent to invoke Article 112 in the Final Act, in order to protect its real estate market, it subsequently cast its net much wider in its own domestic legislation.

In Act No 34/1991 on "Investment by Non-residents in Business Enterprises", as amended by Act No. 121 of 27 December 1993 and Act No. 46 of 22 May 1996 – in Article 12 - is the provision that allows the Minister of Commerce to block a particular foreign investment if he "considers it threatens national security, public order, public safety or public health or in the event of serious economic, social or environmental difficulties in particular economic sectors or particular areas which are likely to be of a lasting nature".

In the case of the investment of a resident in a member state of the European Economic Area, it states, "the provisions of Articles 112 and 113 of the Agreement on the European Economic Area shall be observed". Interestingly, though, in response to the 2008 financial crisis, the Icelandic government did not use Article 112 but instead invoked Article 43 to restrict the free movement of capital.

The use of Article 112 has not been confined to just these four countries – or indeed any specific Member State. On 15 December 1995, via Regulation No 2907/95, the Commission invoked the Article on its own account, making the release for free circulation of salmon of Norwegian origin conditional upon observance of a floor price.

It should not be thought, however, that these applications amount to the full extent of the reach of Article 112. The application of the article is entirely dynamic. In the Accession Treaty for Croatia, Article 37 allows for a response to "difficulties arise which are serious and liable to persist in any sector of the economy or which could bring about serious deterioration in the economic situation of a given area", allowing for the application of Article 112 to this provision. Other accession instruments have the same provision.

Transition to the EEA

Given that continued membership of the EEA might afford some flexibility in the application of the principle of free movement, there is a possibility that other members states could block the transition of the UK from EU member of the EEA to Efta member – assuming that the UK was able to rejoin Efta.

So far, we have taken our advice from the Efta Secretariat on this, which takes the view that transitional arrangements are nowhere set out in the EEA Agreement, and will thus have to be settled politically. One possibility is that on leaving the EU, the UK also leaves the EEA and thus, after joining Efta (if we are allowed in), has to apply to rejoin the EEA – this requiring the unanimous agreement of all Parties.

However, on the basis of previous experience, there is an argument for suggesting that the UK can transition from the EU to Efta while remaining in the EEA. The evidence for this rests with the EEA Agreement of 1992, when Austria, Finland, Sweden and Switzerland were also members of Efta, becoming members of the EEA by virtue of their Efta membership.

Then, in 1995, Austria, Finland and Sweden left Efta to join the EU but were not removed from the list of Efta states in the EEA Agreement until 2004. There was, therefore no issue to deal with on transition. Switching the names from one pillar to the other was dealt with as a minor administrative adjustment.

For the UK and Article 50 talks, this has huge implications. It would appear, on the face of it, that membership of the EEA can continue as long as we join Efta. And, if affirmation of this principle is required, it can probably be secured by agreement not with the EU but with the EEA Council by consensus, which does not even require a formal vote.

One might take it that, in view of the positive response from Schultz to the Swiss proposal to introduce unilateral safeguard measures, and the recent statement by French finance minister Michel Sapin, declaring that everything will be on the table in the future talks with the UK, including freedom of movement, there may be some political support for a seamless transition.

Safeguard measures and the United Kingdom

The outcome of a leveraged deal, using Article 112 as the initial platform would – if the Liechtenstein (and potentially the Swiss) solutions prevail – be formal amendment to the EEA Agreement permitting the UK to impose agreed quotas on immigration from EU Member States.

In relation to the solution preferred by some campaigners, the Australian-style points system, a quota system does not immediately answer the requirement, although it could prove an attractive alternative. The crucial issue here is that the "points system" description is a misnomer. Of the migrants admitted to Australia, only 23 percent are afforded entry as a result of points allocation. The overall limit is an arbitrary quota, set annually – currently at 190,000. This is, by any measure, a quota system.

As to the detail, the essential point – it would seem to this author – is that a fully worked-up case must be made for restrictions, using the Article 112 criteria of "serious economic, societal or environmental difficulties", even if the Article itself is not invoked.

In a 1992 proposal for a Council Regulation (EEC) "concerning arrangements for implementing the Agreement on the European Economic Area", procedures were laid down for implementing Article 112. It thus proposed that, where a Member State requested the Commission to apply safeguard measures, "it shall provide the Commission, in support of its request, with the information needed to justify it".

That should provide a sufficient template for the UK in relation to its Brexit negotiations, permitting a reasoned settlement which is capable of attracting political support.


This post sets out possibilities for a solution to the conflict between post-exit participation in the Single Market and limitations on freedom of movement. It shows that, contrary to claims by the Commission and a widespread belief, freedom of movement provisions are negotiable, and that a legal base within the EEA Agreement exists for a settlement.

Although based on Article 112, which acts as a longstop, the expectation would be of a formal amendment to the EEA Agreement, brokered through the EEA Council rather than the European Council, outwith the formal framework of the Article 50 (TEU) negotiations – but linked to them.

Should we chose to invoke Article 112, the important thing to recognise is that it is not bending or twisting the law. Nor is the Article an emergency provision or a "loophole" – it is a fundamental part of the EEA Agreement. Thus, to enlist it to cap immigration is to use it precisely for the purpose for which it was intended.

Given that – for Efta states – its application is unilateral, as an Efta member, the UK would be entitled to invoke it, this being entirely in accordance with the provisions of the treaty, recognised even by the Schuman Foundation.

Without the EEA solution, there is the possibility that there cannot be a resolution to the conflict between those who regard the need to limit immigration from EU Member States as paramount, and those who see an overwhelming requirement to protect participation in the Single Market.

Even then, if there is a negotiated immigration quota, there is the issue of enforcement. It is one thing applying quantitative restrictions. It is quite another enforcing limits in a large country (as opposed to Liechtenstein), where illegal immigrants can melt into their own resident communities and disappear.

Those who hold that we must abolish unrestricted freedom of movement, therefore, need to understand that imposition of controls, per se – enabled by leaving the EU - does not, in itself, bring immigration under control. Enforcement of immigration controls and a substantial raft of other measures will be required.

Additionally, if the initial exit settlement is only an interim measure, adopted for the purpose of easing our rapid exit from the EU, there is an argument for accepting a sub-optimal settlement if no other outcome is available. Once we are no longer members, it will be possible to work on a longer-term settlement which deals more satisfactorily with the freedom of movement provisions.

Crucially, it must be stressed, the bulk of these negotiations are not conducted within the framework of the Article 50 negotiations, but via the EEA Council, which will require some deft legal footwork if its actions are integrated with the UK exit settlement. But all in all, the prospect of a managed compromise on trade and free movement of people, via the EEA Agreement, must be looked at - and urgently.