自用-网页读书记录

READING
1126k words

目录

已读

✅️ The Very Secret Society of Irregular Witches

✅️ Eleanor Oliphant Is Completely Fine

✅️ Law for Dummies

正在读：Unlocking EU Law

1 The historic origins and character of the EU legal order

DOI: 10.4324/9781003218562-1

1.1 The origins of and background to the Treaties

1.1.1 The background to the idea of a single Europe

One problem that appears to confront students of EU law is an apparent assumption in the United Kingdom that Europe is something foreign, that it refers to a place and to people across the English Channel that have nothing to do with the United Kingdom. This is, of course, not the truth, since the United Kingdom is a part of the European continent and also became a member of the European Community (now called the European Union – the use of the term Community/Community law tends to historically refer to the pre-2009 version of the Union) since signing the Treaties and ratifying membership in the European Communities Act 1972. The United Kingdom was a member of the European Community for just under five decades, from 1 January 1973 until 31 January 2020.

The continuation of the United Kingdom’s EU membership was put to a popular referendum twice. First in 1975, in which MPs of whatever party were free to support the campaign to deliver a ‘yes’ or a ‘no’ vote, according to their consciences and their commitment. Following a high turnout (64.4%), Britain chose to retain EU membership by a majority vote of 67.2%. By contrast, there was a 72% turnout in the second popular referendum in 2016, where 52% of the British public voted to leave the European Union.

Both referenda were led by Euro-scepticism, which is not uncommon within the country, and it would not be unreasonable to suggest that this, in part at least, follows the encouragement of certain elements of the media. This Euro-scepticism seems to be founded in two critical misconceptions about the nature and the role of the Community (now the Union since the Treaty on European Union signed at Maastricht in 1992):

  • That the idea of European unity is a new and modern political concept – in fact it is anything but. The idea is certainly as old as Europe itself and has been put into place before but in different political form.
  • That the Community was based mainly on co-operation between the Member States that may or may not be followed. Again, this is not the truth. The Community legal order was:

based on the objectives of the Treaties that have been agreed by all Member States, and it works towards the achievement of the single dominant purpose of the Community: the full economic integration of the Member States towards a truly united Europe.

In fact, it is unusually appropriate that the Treaty responsible for creating and framing the concept of the Community should be referred to commonly as the Treaty of Rome (this is, of course, the second Treaty of Rome, the first being the EURATOM Treaty). It is appropriate because it is possible to see the Roman Empire as the first real attempt to unify Europe. Of course, in this case it was by a colonial power and through military dictatorship. Nevertheless, the whole purpose of the empire was economic, increasing availability of resources and markets and control of trade.

Following the Romans, it is possible to identify a number of situations demonstrating the same aspiration of a united Europe:

  • The original view of Christendom expounded by the Roman Catholic Church, or more precisely of the Papacy that led it, was very much a European ideal, and everything outside Europe was considered to be barbarian.
  • Charlemagne moulded and reigned over the so-called Holy Roman Empire which covered much of modern Europe. This was obviously a political empire and it soon fell apart after his death as it passed to his sons, becoming the principalities that form the basis of many of the modern nation states.
  • Henry IV of France also tried very hard to create a Christian Commonwealth of Europe.
  • In the nineteenth century, Napoleon Bonaparte declared a stated ambition and policy for a European empire.
  • In the twentieth century, Hitler’s expansionist philosophy was also aimed at a Reich that would control most of Europe.
  • Even besides the above, there have been numerous philosophers, including Kant, Rousseau, Marx and Nietzsche, who have all presented a united Europe as an ideal, and this despite them representing the widest spectrum of political philosophy, from Fascism to Communism.

There has always been, in the widest sense, a European identity and the desire for European unity; and historically, the so-called Euro-sceptics were in a minority. However, Euro-scepticism has been on the rise across Europe, and whilst it had a particular foothold in the United Kingdom which previously was seen as marginal, the referendum on the membership of the UK in the EU on 26 June 2016 revealed a deeper-rooted sense of disenfranchisement from the European project.

The Euro-sceptics have always shown hostility towards the concept of a federal Europe, despite the fact that in its foundations the Community looked towards an eventual political unity as well as an economic unity. The British hostility towards federalism is even stranger since the idea of a federal Europe originated in Britain. It was suggested by Winston Churchill (who was Prime Minister during the war years) before the Second World War and supported then because of the fear of a fascist Germany. In fact, Jean Monnet, who is accepted as being the ‘intellectual father’ of the concept of European Union (EU), himself credited the origins of the ideal to the British. Yet today, the UK has exited the EU under the Article 50 TEU procedure, which was introduced by the Treaty of Lisbon and has been used for the very first time. This is a historic phenomenon which has changed not only the nature of the relationship between the UK and the EU, but also has repercussions for the development of the EU as a polity and for the UK as a non-EU country.

Finally, when considering the origins of the European unification concept, it is worth noting that whatever philosophy has ever been put forward for the EU, it has always included the integration of a binding legal order as well as any economic, social or political union.

1.1.2 The origins of the Union as a Community

It must be remembered that the idea of European unity is based on a very noble and worthwhile principle: the avoidance of war in Europe. War between various nation states, most notably France and Germany, had been an almost constant feature for several centuries before the absolute devastation in the two ‘World Wars’ of the twentieth century.

The late nineteenth and early twentieth centuries saw various attempts at European integration of different types with this object in mind. The French ‘Briand Plan’ of 1929–30 is a classic example.

Nevertheless, economic and political imperatives conspired to create even more disruptive conflicts. As a result, following the Second World War there was an even stronger recognition of the need to avoid future conflicts. Certain other key points were also recognised in the immediate post-war period:

  • that the Treaty of Versailles, while devised to prevent German hostility, had actually been a total failure and had led in part to the rise of Nazi Germany;
  • that the only successful means of preventing war in Europe was not to repress the German state but to tie Germany into a European partnership.

Former British Prime Minister during the war, Churchill in fact encouraged the idea of a form of European unity; and in a speech in Zurich in 1946, shortly after the close of the war, Churchill (by then Leader of the Opposition) suggested:

We must build a kind of United States of Europe. In this way only will hundreds of millions of toilers be able to regain the simple joys and hopes which make life worth living…. The structure of the United States of Europe, if well and truly built, will be such as to make the material strength of a single state less important.

I Ward, A Critical Introduction to European Law (Butterworths, 1996), pp. 6–7

It is worth noting that Churchill was not so much concerned with all nations losing power as he was with Germany losing power.

A European Union of Federalists was then established in 1946. Its supporters on the Continent argued that the way forward was the creation of ‘supranational’ bodies. In the Montreux Resolution of 1947 the group suggested:

A variety of intergovernmental conferences quickly followed, at which agreements were reached to set up new organisations on either a global or merely a European scale. These organisations included:

  • the International Monetary Fund (IMF);
  • the General Agreement on Tariffs and Trade (GATT);
  • the Organisation for European Economic Co-operation (OEEC) (this in effect was the creation of the United States through the ‘Marshall Plan’ of 1947 – the idea being to create an organisation that would administer the financial aid that was needed to repair the damage done to Europe during the Second World War);
  • the Council of Europe (this was a most significant development for its creation of a European Convention on Human Rights and a European Court of Human Rights);
  • the BENELUX Union (a union between the countries of Belgium, the Netherlands and Luxembourg – still among the smallest nations in the EU – significantly, though, the origins of supranationalism can be seen in this organisation for economic co-operation).

If the institutional origins of European Unity began with the Marshall Plan, then the next major phase was the ‘Schuman Plan’. Robert Schuman was the French Foreign Minister, and Jean Monnet was another French politician with responsibility for economic planning who actually drafted the plan for Schuman. Both men were strong supporters of the idea of European co-operation.

The basis of the plan was the integration of French and German coal and steel production under the control of what was referred to as a ‘higher authority’. This is the narrow agenda of the plan. The broader agenda was that it should represent the first stage towards a federal state of Europe.

The plan then led to and formed the basis of the first Treaty: the European Coal and Steel Community Treaty, formally signed as the Treaty of Paris in 1951. A significant point concerning all the various ‘plans’ for a united Europe is the concern to incorporate a legal order as well as economic and political union. This was demonstrated in the Treaty of Paris which created a Community with supranational institutions.

1.1.3 The creation of the Treaties

As a result of the Schuman Plan, the first of the Treaties creating the European Community, later to develop into the EU, was the European Coal and Steel Community Treaty (ECSC Treaty) – the Treaty of Paris 1951.

Focusing on the production of steel and coal was not a random choice, since they were primary engines of war. Both France and Germany had regions where steel and coal were major factors in the economy; therefore as a potential cause of conflict between the two countries, establishing close economic co-operation in these specific fields was a logical means of achieving peace. Indeed, the aim of the ECSC Treaty was to prevent further hostilities within Europe and to establish stable relations between countries that had formerly been at war. Not surprisingly, Monnet was made the first President of the ECSC.

The Treaty in essence devised the institutional framework of all the Communities, with the creation of key political institutions: a ‘High Authority’ (later to be the Commission) to act as the executive arm, controlling production; a political assembly of representative members from the various Member States; and a Special Council of Ministers, with a partly legislative and partly consultative role. Of course, since it was always envisaged that economic integration must be backed by a legal order, a Court of Justice was also added later.

The Treaty was initially signed by six European nations:

  • France
  • Germany
  • Italy
  • Belgium
  • the Netherlands and
  • Luxembourg.

The last three were already closely linked through the BENELUX Union. The UK significantly declined the opportunity to join the Community. Britain had suffered a lot less devastation than many of the mainland European nations, and political thinking of the time leaned towards the so-called special relationship with the United States of America as having more to offer than membership of an economically integrated Europe, and it saw its main market as the Commonwealth.

Nevertheless, economic integration was established with the Treaty, although there were still setbacks in terms of achieving unity. For instance, another initiative aimed at preventing conflict, the European Defence Community, collapsed in 1953. Still, the move towards greater economic integration continued. The Spaak Committee Report of 1952 recommended the placing of atomic energy under a single authority and also called for the creation of a ‘Common Market’:

The report led directly to the creation of two more Treaties. These were the two Treaties of Rome that were signed by all six members of the ECSC on the same day in 1957, the European Atomic Energy Community Treaty (EURATOM) and the European Economic Community Treaty (later to be re-named the EC Treaty).

From the start, the three Communities shared the Common Assembly and also the European Court of Justice (ECJ). Initially the High Authority of the ECSC and the two Commissions of the other two Communities remained separate.

1.2 The basic aims and objectives of EC law and the concept of supranationalism

The major aim of the Treaties was economic integration, but the Treaties also reflect the wider purpose behind that integration. In the preamble to the EC Treaty, it is clearly stated that the central objective of the signatories to the Treaties was to ‘lay the foundations of an ever closer union among the peoples of Europe [by] pooling their resources to preserve and strengthen peace and liberty’.

The union was to be achieved by integration of the Member States’ economic and monetary policies for the creation of a Common Market free from internal barriers to trade. The objectives of the EC in seeking to attain this Common Market were quite simply stated in the EC Treaty itself, later to be amended and built on by the subsequent Treaties:

  • to promote throughout the Community a harmonious development of economic activities;
  • a continuous, balanced expansion;
  • an increase in stability;
  • an accelerated raising of the standard of living and quality of life and closer relations between the states belonging to it and sustainable development of economic activities;
  • a high level of employment and of social protection;
  • equality between men and women;
  • sustainable and non-inflationary growth;
  • a high degree of competitiveness and convergence of economic performance;
  • a high level of protection and of improvement of the quality of the environment;
  • economic and social cohesion among the Member States.

The objectives dictated the structure of the Communities that were created in the Treaties. Key features of the structure can be identified in the EC Treaty:

  • the introduction of four basic freedoms – freedom of movement of workers, the right to establish a trade or profession and to provide services freely, the free movement of goods, and the free movement of capital within the Community – all with a view to the removal of internal barriers to trade;
  • the progressive approximation of economic policies of the Member States and the harmonisation of national laws in key economic areas such as agriculture, transport and trade;
  • the creation of a Common Customs Tariff to regulate imports into the Community from other countries;
  • the creation of a Common Commercial Policy to regulate trade between Member States and other countries.

Inevitably, different Member States had different things to gain from the Treaties and the creation of the three Communities. As a result, from the start proper integration was often hampered by national self-interest. This ensured that development could only be ‘incremental’ and that, often, principle would be sacrificed to what was achievable.

In fact, even before the creation of the Treaties, there was a division in attitudes between those who retained a full federalist ideal and those who preferred a more co-operative progress towards integration, this latter group often being referred to as the ‘functionalists’:

  • the federalists believed in the creation of a single supranational body that would in effect take over all of the functions of the individual Member States;
  • the functionalists preferred to focus on individual areas of the economy and put these under joint management to improve efficiency and with the view that co-operation would then naturally spread from one economic area to another.

The ECSC was based on the attitudes of the functionalists. This gave way to ‘neofunctionalism’ which is demonstrated in the broader principles of the EC Treaty and which reflects the need to limit the power of national governments to legislate in certain areas. This is the basis of supranationalism.

One of the major problems in achieving integration is in defining the constitution of the Community and how it becomes part of the law of the individual Member States. It is of course possible to argue that the constitution of the EC is represented by what is laid down in the Treaties and the objectives that they contain. However, the EC constitution has the added complication of being founded in Treaty relationships, i.e. international agreement, that are then entered into by sovereign states.

In this way, another problem concerns the way that the Treaties are incorporated into Member States’ law, because of the different natures of the constitutions of the different Member States. These are of two distinct types:

  • Monist constitutions: of the original six members these included both France and the Netherlands. In such constitutions, the Treaty is automatically incorporated into the national legal system at the point of ratification.
  • Dualist constitutions: of the original six these included Germany, Italy and Belgium. In such constitutions, the Treaty integrates the national legal system only after it is transposed into national law. Significantly, this was the case for the UK, where the Treaty could only be incorporated into English law after enactment by an Act of Parliament, i.e. the European Communities Act 1972.

As a result of the different methods of incorporation, there can be a wide variance in how the Treaties are then interpreted and applied in the individual Member States. Consequently, in order for the legal order of the Community to have any effect, the Community institutions must be ‘supranational’ in character. In other words, in relation to those areas that are covered by Treaty obligations, the institutions and those Community obligations must take precedence over national law and institutions.

As a result of this, the ECJ has proved to be one of the most significant institutions both in administering and defining EC law, and also in developing the principle of ‘supranationalism’. Both have been vital in ensuring the universal application of the Treaties.

Commentators have frequently accepted the importance of the Court in furthering the objectives laid down in the Treaties. ‘To avoid disparities arising out of different national approaches to the incorporation of EC law and to ensure uniformity in its application, the Court of Justice has developed its own jurisprudence on the Supremacy of EC law’ (P Kent, Law of the European Union (Longman, 2001, p. 55).

ECJ has uniformly and consistently been the most effective integration institution in the Community. Its role was established in A220 [Article 164]: ‘The Court shall ensure that in the interpretation and application of this Treaty the law is observed.’ From its very inception in the Treaty, the ECJ set about establishing its hierarchical authority as the ultimate court of constitutional review. In this area two areas in particular are important. First there is the role of the ECJ in controlling member state courts, and, second, there is the role of the Court in managing the incessant inter-institutional struggles.

I Ward, A Critical Introduction to European Law (Butterworths, 1996), p. 52

1.3 The development of the Treaties (from the European Coal and Steel Community Treaty to the Treaty of Nice)

During the early years of the Community, between 1958 and 1965, there was some successful progress towards economic integration. The period in any case was one of economic boom. The creation of a ‘Customs union’ was underway, with the removal of tariffs. However, there was much less progress towards the creation of an actual Common Market. Some competition policy was put in place and there was some movement made to achieve the free movement of workers. Nevertheless, there was perhaps over prominence of the Common Agricultural Policy (CAP). The ECJ did, of course, do much even in these early years to define the character of the legal order.

Nevertheless, the Common Market should have been achieved within the so-called transitional period referred to throughout the EC Treaty. This comprised three four-year periods or a total of 12 years from 1957, so the Common Market should have been in place by 1969.

However, in the 1960s even the neo-functionalists faced problems with their proposed method of achieving integration. Much of this was the fault of France and in particular President De Gaulle. In 1965–66 De Gaulle created a crisis over qualified majority voting, with France withdrawing from participation in the Council. This was partly resolved by the Luxembourg Accord, which effectively created a sort of veto for the Member States on major issues. This compromise created what has been described as the intergovernmentalist approach. In effect, what it did was to shift more power to the council and away from the supranational bodies.

In any case the years between then and 1986 and the adoption of the Single European Act are commonly characterised as years of stagnation in terms of integration. It is possible to explain the lack of development during this time by the economic recession and also the retreat into national interest. This could particularly be said of the UK from 1979 to 1997.

There were, of course, other areas in which the Community did develop:

  • by enlargement of the number of Member States;
  • by a broadening of policy both inside and outside the original Treaty objectives;
  • by the judicial activism of the ECJ (specifically in the development of the concept of supremacy, and by the development of the means of securing enforcement of the objectives of the Treaty through the processes of direct effect and its associated alternatives).

Another major development has been the creation of further Treaties developing and expanding the original Treaties. To start with, the Merger Treaty of 1965 established a single set of institutions to preside over all three Communities.

1.3.1 The Single European Act

The Single European Act (SEA) of 1986 represents the effective ‘re-launch’ of the Community. A Commission White Paper of 1985 identified that there were a number of pressing problems facing the Community. One of the most critical of these was the fact that the Single Market had not yet been achieved even 16 years after it should have been in place. The Commission therefore proposed to set a new deadline for achieving this aim and to create the means of making it more attractive to the Member States. It was able to push forward for four main reasons:

  • the dynamism of the then Commission President, Jacques Delors;
  • a commitment to further integration by key political figures within the Member States themselves, the most important among these being Chancellor Kohl of Germany and President Mitterrand of France;
  • the relative economic prosperity at the time of many of the Member States, which meant that they also had fewer urgent national problems to deflect them from their Community objectives;
  • a growing realisation in any case that survival depended on an effective EC.

The result was the Single European Act 1986. This set a new time scale for the implementation of the Single Market of 31 December 1992.

Besides refocusing on the principal aim of the original Treaty, the SEA also included constitutional and institutional reform:

  • a new law-making process was devised – the co-operation procedure – that gave Parliament much greater involvement in certain measures and aimed to reduce the ‘democratic deficit’;
  • the European Council was formally recognised and this provided for regular twice-yearly meetings of the heads of state;
  • a new Court of First Instance (CFI) was created to support the ECJ through its massively increased workload;
  • procedural changes also meant that there was an increase of qualified majority voting in Council to speed up the decision-making process;
  • there was also a move towards the idea of European political co-operation and Community competency was extended beyond pure economic objectives to include areas such as social policy and environmental policy.

Nevertheless, the SEA still had its critics, who claimed that it had done nothing more than to begin again the process of integration that should have already been achieved and that it did nothing in effect to remove the veto of individual Member States.

1.3.2 The Treaty on European Union

Commonly referred to as the Maastricht Treaty, the Treaty on European Union (TEU) was signed at Maastricht in 1992. The Treaty was concluded following two intergovernmental conferences in 1989, the first concerning the Commission President’s three-stage plan for European Monetary Union (EMU) and the second concerning political union.

The Treaty itself took a further year to gain force because of difficulties in the ratification process in certain of the Member States and objections to the federalist character of the word ‘union’.

While the Treaty fell short of what many Member States wanted, it did include some major developments:

  • it did create the Union by identifying all the states comprising the Community as being part also of the Union;
  • it also created the concept of European citizenship, although, unlike those rights enjoyed by workers, citizenship did not fall within the legal order;
  • some constitutional and institutional changes were achieved with the creation of a new co-decision procedure for legislating and the inclusion of a Parliamentary Ombudsman and a Committee of the Regions;
  • the objectives of the Community were amended to include EMU, environmental protection and social policy, and the activities were to include research and technical development, trans-European networks, health, education, culture, consumer protection, energy, civil protection and tourism;
  • it also created the ‘three-pillar’ structure of the EU (which has now disappeared since the reforms brought on by the Treaty of Lisbon): added to the central pillar of the Communities would be a second pillar to cover co-operation in foreign policy and security, and a third pillar to cover co-operation in justice and home affairs.

At Maastricht, 11 of the then 12 Member States also agreed on the Social Charter but the UK opted out, thus necessitating the Protocol procedure.

1.3.3 The Treaty of Amsterdam 1997

The Treaty of Amsterdam 1997 (ToA) was the result of intergovernmental conferences in 1996 and 1997. It was agreed in 1997 but did not gain force for another two years. One of the principal problems facing the EU at the time of the Treaty was future enlargement, but in fact the Treaty did very little on this. It is possibly seen as not being so wide-ranging as the TEU but there were a number of key developments:

  • It broadened the objectives of the EU from pure economic aims to the inclusion of some social and political objectives. For instance, where the original Treaty prohibited discrimination based on nationality and Article 141 provides for limited controls in relation to sex discrimination, the Treaty of Amsterdam gave Council the power to introduce laws to prevent discrimination based on racial or ethnic origin, religion or belief, disability, age and sexual orientation. Fundamental human rights are also given greater protection under the Treaty of Amsterdam.
  • The then UK Labour government had accepted the Social Policy Agreement, so the Protocol on Social Policy from the TEU giving the UK an opt-out was repealed in the Treaty of Amsterdam.
  • The Treaty repealed many of the Articles of the original Treaty that were now obsolete and it also grouped together the Articles as amended by the TEU and renumbered them.
  • Institutionally, there was some reform, with the co-decision procedure extended in scope so that only EMU matters fell under the co-operation procedure.
  • There were also some changes to the pillar structure created in the TEU. Under the second pillar, on common foreign policy and security, the Council was given the power to make certain international agreements on behalf of the Member States. Under the third pillar many provisions were moved under the main pillar and the name of the pillar was changed to ‘Police and Judicial Cooperation in Criminal Matters’.
  • A concept of close co-operation was also introduced which allows Member States to co-operate and use the institutions and procedures on areas which are not yet the subject of legislation. This is aimed at reducing some of the tension that exists in certain policy areas.

1.3.4 The Treaty of Nice 2000

The Treaty of Nice (ToN) was agreed in December 2000 but did not gain force until 1 February 2003 because of difficulties in ratifying the Treaty in Ireland, where a referendum initially voted against ratification.

One of the major purposes of the Nice summit was to prepare the necessary Treaty amendments and institutional changes to deal with enlargement of the Union.

  • Institutional changes included changes to the composition of the institutions, with a new allocation of seats in Parliament, an extension to the number of Commissioners and judges in the ECJ.
  • They also incorporated changes to the voting procedures, with the extension of qualified majority voting in Council to account for a new 25-state Union.
  • There was also change to the judicial system, with the creation of a ‘Chamber’ attached to the CFI.
  • The Treaty also approved the Charter of Fundamental Human Rights, with provision for suspension of voting rights where there are breaches.
  • Besides this, measures were introduced for the provision of ‘enhanced co-operation’.

This is the basic idea that some states can be at different levels of integration at the same time by agreeing on certain areas but opting out of others. An example of this would be both Greece and Denmark opting out of EMU. It is sometimes also referred to as ‘variable geometry’ or as ‘multi-speed Europe’.

1.4 Enlargement

Enlargement of the EU has occurred and continues to occur in two ways, sometimes referred to as ‘widening’ and ‘deepening’.

‘Deepening’ refers to the increasing scope of influence of the EU over its Member States. As has been seen above, the extension to the three-pillar structure of the EU widened its influence. There has been a gradual and ongoing move, through the succession of Treaties, towards a social and political agenda well beyond the limited economic context of the original Treaties. This is evidenced by the widening and redistribution of the power of the different institutions and the creation of extra institutions over time. Besides this, the ECJ had steadily defined the scope and effect of EC law.

The context of the word ‘enlargement’ has been characterised by a widening of the EU by territory; which grew significantly since the original Treaties and its original six signatories, i.e. Germany, France, Italy and the three BENELUX countries:

  • in 1973 the six became nine, with the addition of the UK, Ireland and Denmark;
  • in 1981 Greece also joined, making for a Community of ten countries;
  • in 1986 Portugal and Spain also joined, to make it 12;
  • in 1995 the EU was enlarged to a group of 15 with the accession of Austria, Finland and Sweden;
  • 2004 saw the biggest single expansion, with ten nations joining: the Czech Republic, Hungary, Poland, Slovakia, Slovenia, Estonia, Lithuania, Latvia, Malta and Cyprus;
  • in 2007, Romania and Bulgaria joined the EU;
  • in 2013, Croatia became the 28th member state of the EU;
  • since 2020, the Union counts 27 member states after the exit of the United Kingdom.

As of 2023, ten European countries are in various stages of negotiation for entry into the EU, including candidate countries Albania, Bosnia and Herzegovina, Moldova, Montenegro, North Macedonia, Serbia, Turkey, Ukraine, and potential candidate countries Georgia and Kosovo.

Whilst candidate countries are in the process of integrating EU legislation into national law to align with the bloc, potential candidates countries are yet to complete the requirement process for EU membership.

The EU is already the largest trading bloc in the world, and the population covered by the Union is close to 450 million. Gaining membership is not as simple as just making an application.

Before a country can be considered for membership it must now meet three specific requirements:

  • it must show that its institutions are democratic and guarantee the rule of law, protect human rights, and specifically is able to protect minorities;
  • it must show that it operates a market economy that is capable of surviving within the Single Market;
  • it must show that it is able to apply the laws of the EU and can meet the aims of economic, political and monetary union.

One significant point to make about the Union, when considering it in the context of one of the central aims of the devisers of the original Treaties, is that peace within the Member States has been maintained throughout that period.

A map of Europe represents the history of Europe from 400 Before Christ to the present. The regions of the country, such as France, Germany, the United Kingdom, Greece, Spain, Poland, Sweden, etc., are shaded in different colours.

Figure 1.1 Member States of the European Union

1.5 The EU Reform Treaty

1.5.1 The Failed Constitutional Treaty

Later Treaties, as well as the intergovernmental conferences, focused on the difficulties of expanding the EU and the impact that had on the institutions and the administration of the law. A year after the Nice Treaty was signed, at a meeting of the European Council in Laeken, a ‘Declaration of the Future of the European Union’ was adopted.

The Laeken Declaration committed the Union to becoming more democratic, more transparent and more effective, and therefore recognised the need to create a Constitution. The meeting also recognised that the traditional method for reviewing and developing the Treaties, through intergovernmental conferences of the heads of state, lacked transparency and democracy. A decision was taken to convene a Convention of representatives of governments of all the Member States and of the countries which had applied for membership as well as representatives from the institutions and the various committees – a Convention of 105 members in total.

This Convention looked into the various problems associated with taking the Union forward and eventually prepared a Draft Constitution which was presented to the intergovernmental conference at Thessaloniki in June 2003. The Draft Constitution was subsequently agreed by the Member State governments, identified as the EU Constitution and was referred to the Member States for ratification. A number of Member States ratified the EU Constitution by parliamentary vote. Certain Member States, including founder members such as France, decided to ratify by a national referendum. The rejection of the EU Constitution in the French referendum in May 2005, followed by the Netherlands’ rejection in June 2005, created doubts as to whether the EU could proceed with the EU Constitution at all.

One of the main purpose of the EU Constitution was to put into a single document all of the existing Treaties so that it should be readable and clear. The Constitution itself was made up of four separate parts:

  • The first part contained all of the provisions that related to the objectives, powers and decision-making processes of the Union and those relating to its institutions.
  • The second part incorporated the Charter of Fundamental Rights agreed upon at Nice.
  • The third part focused on the policies of the Union and incorporated many of the provisions already contained in the existing Treaties.
  • The fourth part, as usual, was referred to as the Final Clauses and in this instance contained the procedures for adopting and reviewing the Constitution.

1.5.2 The Reform Treaty (the Lisbon Treaty)

Following the effective collapse of the Constitutional Treaty, the Brussels European Council in 2007 decided that the relative inertia resulting from this collapse, as well as the need to develop the institutional reforms begun in the Treaty of Nice 2001, meant that the position needed to be resolved and so the Treaty of Lisbon, the so-called Reform Treaty, was also drafted in 2007.

While arguments raged in UK politics on the question of whether the Lisbon Treaty is just the Constitutional Treaty in a different guise, there are significant differences, although many important aspects of the Treaty remain unchanged.

One of the most important changes for students of EU law is in the different form and structure of the two Treaties:

  • The Constitutional Treaty proposed in effect to create a single written constitution, incorporating but repealing all previous Treaties. Unfortunately, to overcome the objections to the Constitutional Treaty, the Lisbon Treaty follows the format of previous amending Treaties, keeping the other Treaties in place while amending them. As a result, significant cross-referencing is still required between all the various Treaties and the Lisbon Treaty.
  • The proposed change to EU laws and EU framework laws in the Constitutional Treaty is not repeated in the Lisbon Treaty and the traditional legislation, regulations, directives and decisions remain.
  • The Charter of Fundamental Rights would have been incorporated in the Constitutional Treaty; instead, under the Lisbon Treaty, Article 6 TEU was amended to provide only that the Charter of Fundamental Rights has the same legal value as the Treaties, to which it is annexed. However, Protocol 30 allows that, in the UK and Poland (and since 2009 in the Czech Republic), the Charter will create no justiciable rights beyond those already in existence within the national law of those two states.
  • Institutional reforms remain largely intact. Qualified majority voting is extended to other areas, as in the Constitutional Treaty. The introduction of Double Majority Voting remains, although its full introduction was delayed up to March 2017. Parliament was given increased powers, as in the Constitutional Treaty, with the extension of the co-decision procedure (now known as the ordinary legislative procedure) to a wider range of areas.
  • Where the Constitutional Treaty proposed to abandon the rather complex three-pillar structure of the TEU, and place foreign affairs and justice and home affairs within the constitution, the existing structure in essence is to be retained.
  • One contentious issue of the Constitutional Treaty, the creation of a flag, anthem and motto, has been abandoned in the Lisbon Treaty.

Originally the Treaty was rejected in an Irish referendum. A further successful referendum led to the Treaty being ratified by Ireland in 2009, and Poland also ratified the Treaty, so that it is now in force.

1.5.3 The Charter of Fundamental Rights of the European Union

Purpose and content

At the Cologne European Council in June 1999, the heads of states and government of the EU Member States decided that the system of fundamental rights protection, developed under the doctrine of general principles of EU law, should be consolidated in a charter of rights in order to make the latter more visible to the citizens of the EU. Despite the fact that fundamental rights were not considered or included in the original founding Treaties, the European Council nevertheless stressed that they constitute a ‘founding principle of the Union’. Therefore, the use of the term ‘consolidate’ is a key operative word which characterises the nature of the Charter of Fundamental Rights (CFR). The Convention that was charged with drafting the CFR included representatives of the Member States, the European Commission and Parliament, as well as representatives of the Member States’ national parliament and civil society. The Convention did not technically create any new rights, and instead catalogued in a single document all pre-existing fundamental rights which form part of the Union acquis communautaire (Union body of law). The European Council qualified fundamental rights protection as ‘an indispensable prerequisite to the Union’s legitimacy’, therefore the CFR makes these existing rights accessible and visible for the citizens of the EU.

The CFR encompasses 50 rights, freedoms and principles across six chapters on Dignity, Freedoms, Equality, Solidarity, Citizens’ Rights and Justice. The final Chapter VII relates to the interpretation and scope of application of the CFR.

The range of rights covered in the CFR is wider and more inclusive than those of the European Convention on Human Rights (ECHR) and other international instruments. Indeed, the CFR brings together an amalgamation of socio-economic, cultural and moral freedoms, along with civil and political rights in one document, which is a remarkably advanced consolidation of traditional, modern and aspiring rights and freedoms.

Scope of application

From its initial proclamation at the Nice Council in 2000 (which merely made the Charter ‘politically binding’ on the institutions and Member States of the Union), the CFR was eventually given legal force by the Treaty of Lisbon.

Article 6(1) TEU ensures that the CFR is primary legislation. Although annexed to the Treaties in a declaration, the CFR has nevertheless the same constitutional value as the TEU and the TFEU in the EU legal order, as recognised by the CJEU in Kücükdeveci (Case C–555/07).

Following Article 6(1)(3) TEU, the Charter comes with a full package of mandatory interpretative assistance, including the horizontal provisions of Title VII and the Convention Praesidium’s explanations of the Charter’s rights, principles and details of their sources.

On the scope of application of the CFR, the ECJ has held that, with regards to Article 51, the application of EU law implies the application of the CFR. Therefore, where national legislation falls within the scope of EU law, provisions of the CFR will be applicable (Åkerberg Fransson (Case C–617/10)). The ECJ adds that under Article 53 national courts may apply ‘standards of protection of fundamental rights, provided that the level of protection provided for by the Charter, as interpreted by the Court, and the primacy, unity and effectiveness of EU law are not thereby compromised’ (Melloni (Case C–399/11)). The ECJ’s interpretation of the scope of application of the CFR seems closely linked to ensuring the primacy and effectiveness of EU law, as seen in Kadi (Joined Cases C–402/5 P and C–415/05 P) in which the ECJ held that the CFR cannot be overridden by a Member State’s obligation under an international agreement where the application of the latter would contravene a provision of the CFR.

Finally, Article 1(1) of Protocol 30, which also formerly applied to the UK, states that the ECJ has no competence to scrutinise Poland’s domestic courts conformity with Charter provisions. However, the ECJ has taken a rather minimalist approach to the interpretation of this so-called opt-out Protocol by stating in N.S. (Joined Cases C–411/10 and C–493/10) that the protocol does not exempt the UK (when still a member state) and Poland, or their respective domestic courts from complying with the CFR provisions.

In effect, both the European Parliament and the ECJ appear keen that the opt-out Protocol remain a paper tiger with little chance of effectively curtailing the application of the CFR. In a majority vote of 574 to 82, the EP eventually rejected a proposal to amend Protocol 30 that would have added the Czech Republic as a signatory (Official Journal of the European Union 2016/C 055/33). The EP argued that “the function of the Charter is to increase the prominence of fundamental rights and to make them more visible”; and that Protocol 30, if it were interpreted as limiting the Charter, would give rise to legal uncertainty and political confusion, potentially adversely affecting the effectiveness of EU law protection of fundamental rights.

In light of the above, it is apparent that the CFR has become a substantial instrument in the evolving politico-legal nature of the Union and the expanding reach of its legal order.

1.5.4 Post Treaty of Lisbon: from Community to Union

Prior to the ratification of the Lisbon Treaty it was important to understand the different concepts of Community and Union because the European edifice was built on a three-pillar structure. From the foundation of the original Treaties to the time of the TEU, there was a Community; first the EEC and later the EC. The choice of the term ‘community’ by the founding fathers of the EU may have been used as an alternative to the f-word in order to assuage any concerns from the anti-federalist section. ‘Community’ is also a denomination that had no prior political definition and so could be distinguished from the more traditional types of international organisations. Therefore, the neutral characteristic of the term ‘Community’ was ideal for the new European venture of peace-building and integration through co-operation and solidarity between Member States.

More concretely, the initial founding Treaties comprised not only the (economic) objectives of the Community but also the blueprint for its legal (and public) order. Following the TEU, which also established EU citizenship, the more political incarnation of the EU was created. Between the TEU (1992) and the Lisbon Treaty (2009) there was both the EC and the EU at the same time (under the now defunct three-pillar system). In very general terms, the EC was founded by the Treaties, as amended, and was based on law; whilst the EU, stemming from the agreements made at Maastricht, was based on co-operation. This was a cumbersome and unsatisfactory structure which was reformed by the Treaty of Lisbon. Today, the EU is one unitary entity with legal personality, and is based on law (with some exceptions, such as for the CSFP). Whilst it may be helpful, in studying the law of the EU, to understand the difference between the EC and the EU, for this textbook’s purpose to introduce and simplify EU law, suffice to know that any terminology related to Community and EC law refers to the EU pre-Treaty of Lisbon.

1.6 The future

1.6.1 Enlargement – nationally

Although the EU underwent its biggest expansion in May 2004, the Union is not yet at ‘capacity’. Indeed, any nation satisfying the basic admission criteria for membership is entitled to apply for membership.

The process to EU accession can broadly be described as involving three stages:

  1. Candidacy application, including the status of ‘potential candidate country’ and ‘candidate country’.
  2. Accession negotiations, which involves the candidate country going through a process of aligning its national laws with EU law, including integrating EU legislation within its national legal system. There are 35 chapters to negotiate, complete and close, i.e. a policy domain such as taxation or human rights.
  3. Accession treaty, that is an accession agreement between the candidate country and the EU, must be signed and then ratified by all the member states of the EU before the candidate country can fully accede to and become a member state of the EU.

The last country to become a new Member State was in 2013, and to date eight more are at different stages in the process of application to become Member States:

  • Albania – first made a formal application in 2009, although it did not become a candidate until 2014, after the opinion delivered by the Commission in 2010 identifying that Albania needed to deliver on 12 key priorities before accession negotiations could begin. After an EU Commission unconditional recommendation in 2018, the General Affairs Council decided to open accession negotiations in 2020, and the Commission started the screening process in 2022.
  • Bosnia and Herzegovina – was identified as a potential candidate country in 2003, and after several visa and trade alignment agreements, including the Stabilisation and Association Agreement (2015) with the EU, Bosnia and Herzegovina applied for EU membership in 2016. In 2019, the EU Council approved the Commission’s Opinion identifying 14 key priorities for reform before the country could accede to the EU, including democracy, the rule of law and fundamental rights.
  • Moldova – applied for EU membership in March 2022 and became a candidate country by June 2022.
  • Montenegro – another part of the former Yugoslavia, which declared its independence from the post-Yugoslavia state of Serbia and Montenegro in 2006 and applied for membership in 2008. The Commission issued a favourable opinion on its application in 2010 subject to fulfilment of seven key priorities, and it was given candidate status with accession negotiations beginning in 2012. Montenegro also enjoys visa-free entry into Schengen agreement states.
  • North Macedonia – first applied for membership in 2004 and was given candidate country status in 2005. The EU Commission unconditionally recommended accession negotiations in 2009, although in 2015–16 accession negotiations became conditional upon the implementation of the Pržino agreement (a political agreement between the main Macedonian parties to put the interest of the country first and end the national political crisis). Following substantial progress, accession negotiations were opened by the General Affairs Council in 2020, and in 2022 the Commission started the screening process.
  • Serbia – first applied in 2009 and was granted candidate status in 2012. Accession negotiations formally started in 2014.
  • Turkey – the country which has undergone the most complicated and protracted application process ever, applied to join the EU in 1987 and received candidate country status in 1999. However, whilst membership negotiations opened in 2005, they were stalled in 2006 over free movement goods restrictions, compounded by concerns about Turkey’s human rights record and what has been termed ‘democratic backsliding’ in the press. To date, the relations between Turkey and the EU is at an impasse, and a different political route is being explored since the European Council meeting of 19 October 2017.
  • Ukraine – applied for EU membership in February 2022 and obtained candidate status by June 2022.

There are also currently two countries which are potential applicants or potential candidate countries, i.e. they have not yet fulfilled the accession criteria necessary:

  • Georgia – applied for EU membership in March 2022 and was given a European perspective to become a member, i.e. it is recognised as a potential candidate provided it addresses the EU Commission’s key priorities. These include addressing issues of political polarisation; the full functioning, independence and democratic character of all state institutions; judicial reforms for transparency, and more.
  • Kosovo – since 1999, the EU has played an enormous role in the reconstruction and redevelopment of Kosovo through the EULEX rule of law mission, in assisting towards independence and in helping to create a constitution. In 2014, the EU and Kosovo initiated the Stabilisation and Association Agreement, which was signed in 2016, and was followed by the Commission’s Economic & Investment Plan for the Balkans in 2020.

1.6.2 Enlargement – scope of legislation

Over the last 20 years, the scope of EU legislation has expanded significantly, taking in new topics such as discrimination on grounds of sexual orientation and race (discussed in Chapter 18 of this book).

1.6.3 Brexit, disintegration and differentiated integration

In 2016, the British government held a referendum on the United Kingdom’s membership of the EU. After a highly politicised and controversial debate (often lacking in factual accuracy or convincing arguments) between the Leave and Remain campaigns, 52% of the British people voted to withdraw from the EU. The submission of Prime Minister Theresa May’s letter to the President of the European Council on 29 March 2017 under Article 50 TEU triggered the two-year process of negotiation between the institutions of the EU and the UK for its withdrawal from the European bloc, which should have initially become effective on 29 March 2019 (European time).

The negotiations between the UK and the EU were far from straightforward. They substantially hampered by the British government’s lack of clear withdrawal strategy and forward planning on future UK–EU relationship plans. Consequently, the UK had to seek permission from the EU Council to extend the withdrawal twice.

The European Union (Withdrawal Agreement) Act 2020 eventually received Royal Assent on 23 January 2020; and the UK officially exited the EU on 31 January 2020, entering a transition period that ended on 31 December 2020.

The very first withdrawal of a Member State from the Union construct has generated much political and academic debate, as well as soul-searching (see section 1.6.4), about the future of the EU. Concepts of disintegration and differentiated integration have surfaced. Although there is no wide consensus on the meaning and scope of each notion, disintegration can be defined as the process whereby one or more aspects of European integration may – to some extent – be reversed (e.g. no further deepening, broadening or widening of Union competences, policy fields or enlargement). On the other hand, differentiated integration alludes to options for multiple speeds in the integration process, allowing Member States to not participate in certain policy fields, or policy fields to offer various extents of involvement.

These are questions that form part of the wider reflection on the future of Europe – what kind of Union can we expect in the coming decades?

1.6.4 Towards full union …?

For many years, the EU stayed the course on its way to achieving full economic and monetary union, with the single European currency, the euro, successfully launched in 1999 (for electronic transactions) and in 2002 (for all other transactions), and the establishment of the European Central Bank. The next logical step would be for those participating states to hand over further control of their economic policies to a central European authority. This could in turn lead to the next step in integration terms – a full political union – which would probably involve conferring far greater powers on the Council and/or European Parliament at the expense of national Parliaments, especially in the fields of foreign policy and security. However, the 2008 economic crisis and Brexit have both altered this trajectory and forced the EU to start reflecting on its future.

Without the above-mentioned events, the next step in integration terms would have been the ultimate step – full union – which would entail the complete unification of all national economies. That would mean the end of such fiercely protected national policies as the setting of rates of income tax and, with it, national identities. The EU would then become a state in its own right, albeit a federal state with a certain amount of autonomy devolved to the former Member States, which would be re-classified as ‘regions’ of the EU, in accordance with the principle of subsidiarity. A name change would then be necessary, the obvious contender being ‘the United States of Europe’.

Significantly, one of the events that have made closer political integration more likely has been the so-called Eurozone crisis. Economic monetary union was always a necessary step along the road to a true Single Market. However, it proved controversial and unpopular with a number of Member States including the UK.

Nineteen out of twenty-eight Member States in fact joined a single currency. Because of the problems of differences in economic stability of different Member States, convergence criteria were created, and it has been argued that certain states were not really in a position to match the criteria. The problems resulting from the world banking crisis and recession also led into major problems in a number of Member States in the Eurozone. The EU has effectively had to provide financial support to a number of Member States in the Eurozone and as a result has had to intervene also to create greater fiscal discipline in those states, which has led to plans for greater fiscal integration.

In March 2017, on the eve of celebrating the 60th anniversary of the EU, the European Commission published the White Paper on the Future of Europe and the Way Forward. The White Paper presents five different scenarios on the possible avenues in which the Member States might wish to take the EU in the future. The question is whether the Member States wish to further, pause or decrease the European integration process. The five scenarios are as follows:

  • Scenario 1 – Carry on: the EU would pursue its current reform agenda (including on climate and sustainable development issues, financial stability policies, and defence and security strategies) as set out in the European Commission in 2014 and the European Council’s Bratislava declaration of 2016.
  • Scenario 2 – Nothing but the Single Market: the EU would decrease its intervention in wider policy areas beyond the Single Market (e.g. migration, security) and would essentially be an economic union with less regulatory activities.
  • Scenario 3 – Those who want more do more: the EU would allow different speeds to the integration process, facilitating the means for a ‘coalition of the willing’ to work closer with each other in certain policy areas, whilst allowing other Member States to step back or join in as suits them.
  • Scenario 4 – Doing less more efficiently: the EU would revise its key developmental priorities and ring-fence its resources to a selected few – in other words, the EU would do less but better in those chosen policy areas.
  • Scenario 5 – Doing much more together: the EU would decide to deepen the integration process by ‘sharing more power, resources and decision-making across the board’ (a.k.a. go full union).

The aim of the White Paper is to open discussion and debate on the future direction of the EU; something that has arguably been much needed since the creation of the citizenship of the EU by the Maastricht Treaty 1992, i.e. since the EU became a more recognisable (if yet undefined) form of political community. The Nice summit and the Laeken Declaration already alluded to the need for reflecting on the nature of the European venture. The Treaty of Lisbon (as a different incarnation of the failed Constitutional Treaty) failed to answer the Union’s existentialist question. It is Brexit (which was set to inspire a potential Grexit, Frexit or Nexit if national popu-lists are to be believed) that acted as a catalyst for the White Paper’s call to reflect on the EU’s future.

Following the results of the EU Commission’s online public consultation in 2018–19, in which EU citizens were invited to complete a survey on their thoughts on the direction the EU should take, a Conference on the Future of Europe was announced in 2019. The Conference began in earnest in 2021 through a series of European Citizen’s panels. European citizens from across the bloc gathered (in physical or digital meetings) to discuss, debate and share their ideas on many issues pertaining to the future of the Union around nine topics: climate change and the environment; health; a stronger economy, social justice and jobs; the EU in the world; value and rights, rule of law and security; digital transformation; European democracy; migration; and education, culture, youth and sport.

The results of the Conference, presented in a report submitted to the Presidents of all three political supranational institutions (European Parliament, Commission and Council), contains 49 proposals across the nine topics.

SUMMARY

  • There are many examples of former attempts at a united Europe.
  • The EU (formerly the EC) was created following the Second World War with its key aim being: to ‘lay the foundations of an ever closer union among the peoples of Europe [by] pooling their resources to preserve and strengthen peace and liberty’, and this was to be achieved by integration of Member States’ economic and monetary policies for creation of a common market free from internal barriers to trade.
  • The original six Member States has expanded to 27 (following Brexit, the first withdrawal from a member state from the Union).
  • A variety of Treaties has expanded the scope of the original Treaties.
  • The Lisbon Treaty, while an amending treaty, introduces widescale institutional reform and gives legal force to the Charter of Fundamental Rights.

Further reading

Articles

  • Lenaerts, K, ‘Exploring the Limits of the EU Charter of Fundamental Rights’ (2012) 8 EuConst 375–403.
  • Markakis, M, ‘Differentiated Integration and Disintegration in the EU: Brexit, the Eurozone Crisis, and Other Troubles’ (2020) 23(2) J Int Econ Law 489–507.
  • Meyring, B, ‘Intergovernmentalism and Supranationality: Two Stereotypes for a Complex Reality’ (1997) 22 ELR 221.
  • Schramm, L, ‘European Disintegration: A New Feature of EU Politics’ CEPOB (College of Europe Policy Brief) #3.19, May 2019.
  • Wouters, J, ‘Institutional and Constitutional Challenges for the European Union: Some Reflections in the Light of the Treaty of Nice’ (2001) 26 ELR 342.

Books

  • Craig, P and de Burca, G, EU Law: Text, Cases and Materials (OUP, 2020).
  • Douglas-Scott, S, Constitutional Law of the European Union (Longman, 2002), chap. 1.
  • Schütze, R, European Union Law (OUP, 2021).
  • Ward, I, A Critical Introduction to European Law (3rd edn, CUP, 2011), chap. 1.

Website

  • More information on the European Union can be obtained on the website at: http://europa.eu.

2 The political and legal institutions of the European Union

DOI: 10.4324/9781003218562-2

2.1 The origins and development of the institutions

Both the original concept and indeed the major character of the key institutions of the EU actually came about in the ECSC Treaty. The framers of the Treaty realised that in order to have the sort of Community envisaged by the founders, such as Jean Monnet, which could work effectively and prevent outbreak of another European war, it was vital to have both a legal order and a legislative system that could work independently of the Member States.

The next realisation was that this in itself would require the creation of institutions that were demonstrably supranational, bodies that could work independently of the Member States for the achievement of the aims of the Treaty.

In this way the ECSC Treaty introduced the essential concept of ‘community’, as we know it now. The European Coal and Steel Community was created with a number of significant features in mind:

  • that the Community should have a distinct legal personality;
  • that it should work through and be represented by entirely autonomous institutions;
  • that Member States which became signatories to the Treaty should agree to cede (or more precisely pool) certain aspects of their national sovereignty in order to achieve defined objectives identified in the Treaty.

The institutions of the new ‘Community’ were created on this basis. The first of these institutions was originally known as the ‘High Authority’. This body was to effectively control the production of coal and steel and was given power by the Treaty to make legally binding decisions as well as recommendations.

This body was supplemented by the creation of a ‘Common Assembly’. This was a representative assembly to be made up of members from the various signa-tory states. It would have only very limited supervisory powers.

To these bodies was added a ‘Special Council of Ministers’ of the Member States. The Council’s powers were partly legislative and partly consultative.

The final body devised in the Treaty was a ‘Court of Justice’. The straightforward role of the Court was laid out in Article 31 of the Treaty, ‘to ensure that in the interpretation and application of this Treaty … the law is observed’.

When both the EEC Treaty and the EURATOM Treaty were created they followed exactly the same pattern, establishing ‘Communities’ with four key institutions, all autonomous, and all able to exercise powers stipulated in the Treaties creating them.

In these two Treaties, bodies called ‘Commissions’ in fact replaced the ‘High Authority’ of the ECSC. These Commissions initially had more limited powers and ranked behind both the Assembly and the Council.

At the same time that these two later Treaties were signed, a Convention created a single Assembly and a single Court of Justice that would represent all three Communities rather than having separate institutions for each.

Much later, in the Brussels Treaty 1967 (referred to as the Merger Treaty), a single Council was introduced for all three Communities. Also in the same Treaty the High Authority of the ECSC and the two Commissions of the other two Communities were merged into a single Commission. The Communities were nevertheless kept separate and the powers exercised by the individual institutions would be by reference to the specific Treaty.

The institutional structure of the EU is in fact quite unique. For instance, it is not like the governing structure of any international organisation since the institutions are able to exercise sovereign powers that are in effect transferred to them by the Member States on joining.

Neither is it like any normal parliamentary democracy. There is of course no clear separation of powers between the executive and legislative functions as there would be in the constitutions of the Member States. However, there is a separation of interests.

There are complex legislative processes with both the Commission and the Council being potential sources of interpretation for legislation and the possibility of amendment or consultation by Parliament before legislation is in fact passed or created by the Council. Indeed there are even other bodies before which draft legislation might be presented before it is actually passed by the Council.

The nature of the different institutions and the various powers or responsibilities given to them means that the Commission is in fact a sort of executive but one without real executive powers. The Court of Justice is not a court of the type that we are familiar with but exercises a supervisory function and ensures that the division of power between the other institutions and between the EU and the Member States is maintained and preserved.

In keeping with the wishes of the founders of the Treaties the institutions all have broad autonomy and also their own rules of procedure. However, whatever the power of the individual bodies the overriding ruler of the EU is in fact the Treaties themselves. Everything done by the institutions must be for the furtherance of the objectives set in the Treaties.

These four basic institutions are still the main ones but they have also been added to over the years through provisions in the original Treaties or in the different subsequent Treaties:

  • Article 4 of the EC Treaty allowed for the setting up of both an Economic and Social Committee and a Court of Auditors (to audit the accounts of the institutions). These two bodies have been created and the latter was made an independent body as the result of the TEU.
  • Article 4 also allowed for the creation of COREPER, the Committee of Permanent Representatives. This is a vital part of the legislative process.
  • The SEA 1986 then also established the Court of First Instance, a vital addition to ease the burden on the ECJ.
  • A Committee of the Regions was also added by the TEU 1992.
  • The TEU also provided for a European Investment Bank as well as a European Central Bank.

Besides new institutions being added to the original four, the role of the original institutions has also changed over the years. The new institutions have created extra complexity in the overall administration but have not necessarily improved it.

The balance of power between the Council and the Commission has shifted in a way that was never envisaged in the Treaties. As a result, intergovernmental decision-making took precedence over supranational authority. In fact, the power of the Council increased because of the evolution of the European Council, the importance of the presidency and the co-operation procedure (today known as the Ordinary Legislative Procedure). Besides this, the work of COREPER in many ways supports the individual states.

The role of the European Parliament (EP) has also increased. It now has a much greater and more significant role in legislation as a co-legislator with the council of the EU, which makes the European Parliament (EP) a key player in the law-making and decision-making processes in the EU public order. It also has in effect control over a significant part of the budget and in its supervisory capacity was able to react dramatically in forcing the resignation of the Commission in 1999.

2.2 The Council of Ministers and the European Council

2.2.1 The Council

The Council is governed by Articles 237–243 TFEU and Article 16 TEU. While the Treaty itself merely refers to it as ‘Council’ it was traditionally referred to as the ‘Council of Ministers’. However, following the TEU the Council chose to change its name to the ‘Council of the European Union’. The Lisbon Treaty also refers to it as the ‘Council’.

Composition

According to Article 16(6) TEU: ‘The Council shall consist of an authorised representative of each Member State at ministerial level, authorised to commit the government of that State’, although the original intention was that the Council should in effect be a college of delegates with each state appointing a Minister for the specific purpose.

Therefore, the Council, in terms of its membership, is a fluid concept with what amounts to a floating membership. Given that it is composed of Ministers from each Member State, its membership is therefore political in nature (i.e. ministers are politicians rather than civil servants). In this sense it is unlike other bodies with similar functions.

The consequence of the definition is that the specific identity of a Minister attending a meeting of the Council depends entirely on the subject of discussion at the meeting. If the meeting concerns agriculture it will be the Agriculture Minister. If it concerns transport it will be the Transport Minister, and so on. The important point is that each Minister has the power to commit the whole government of the Member State but is then accountable to his national Parliament which is said to ensure democratic legitimacy. The composition of the Council would also have been changed significantly under the Constitutional Treaty.

The role of Council

Article 16(1) TEU identifies the role of Council in the following way: ‘Council shall, jointly with the European Parliament, exercise legislative and budgetary functions. It shall carry out policy-making and co-ordinating functions as laid down in the Treaties.’ In fact, Council actually has six key responsibilities:

  • Legislation – it is often said that ‘the Commission proposes and Council disposes’. The Council, through the different legislative procedures, acts on proposals from the Commission and with the advice of Parliament passes EU law. In this respect Council is the key decision-maker within the EU.
  • Co-ordinating economic policy within the Union – the EU is an economic union and the Common Market is a significant part of that.
  • Forming international agreements on behalf of the Member States – possibly on trade, technology, development, etc.
  • Approving the EU budget – the budget is in two parts: compulsory (for instance agricultural) and non-compulsory (which deals with the upkeep of the institutions). Both Council and Parliament must agree on the budget.
  • Developing common foreign and security policies – this is only based on cooperation as Member States retain their own independent control over, e.g. defence. But there is much to be gained by working together and there is for instance an EU ‘Rapid Reaction Force’ of military personnel.
  • Co-ordinating co-operation of justice systems of the Member States – different Member States have different legal systems but the EU can usefully work in co-operation on areas such as drug-trafficking, international fraud and terrorism.

The presidency

The Council has a rotating presidency which is for a six-month period. Each period is headed by a different Member State.

Indeed, since the Treaty of Lisbon (ToL), the Council presidency operates under the ‘trio’ mechanism or programme, whereby Member States are each called to preside over the Council activities in groups of three. For instance, France, Czech Republic and Sweden were the presidency trio for the following respective periods: January–June 2022, July–December 2022, and January–June 2023. The trios work closely together to ensure the continuity of the EU’s legislative agenda and key areas of policy activities. The main tasks of the Council presidency are to plan and chair Council meetings; and to represent the Council vis-à-vis other EU institutions. Because the Council meets with a different make-up according to the area covered, this allows each configuration to be headed by a representative of each Member State on rotation.

As with the former system there are advantages to being President since the President sets the agenda.

Voting

The Council is one half of the law-making duo of the EU, along with the European Parliament.

Its voting procedure can three forms: simple majority, qualified majority voting (QMV), and unanimous. The first two types are identified in Article 238 TFEU:

  • Simple majority requires at least 14 Member States to vote in favour of a measure.
  • Qualified majority voting (QMV) requires at least 55% of Member States, together representing at least 65% of the EU population, to vote in favour of a measure – and is known as the double majority rule.
  • Unanimous voting is the third type, whereby all Member States have to unanimously votes in favour of a measure. Unanimous voting is required in a number of policy areas deemed sensitive (e.g. since the ‘Luxembourg Accord’, where national interests were at risk; taxation, common foreign and security policy, citizenship), which are exhaustively stipulated in the Treaties.

Unanimous voting and the old system of majority voting offered an opportunity for some countries to veto decisions and therefore stall Union policy development. The Treaty of Nice reduced quite drastically the number of areas requiring unanimity and opened the way for a phasing-in of the qualified majority voting system prevent stagnation and large states from abusing their power at the expense of smaller ones.

Now, QMV is the default voting system of the Council as per Article 16 TEU, and it appears as though the Council is considering the use of ‘passerelle’ clauses in the TEU, in order to extend the areas in which QMV might apply without the introduction of major Treaty amendments – thus effectively reducing the prevalence of national vetoes.

Given the diversity of the countries that are members of the EU, voting was weighted and proportionate, i.e. Member States were given more votes according to the size of their populations and their economic influence. However, the weighting was also artificial in these terms to favour the smaller countries. Under the Lisbon voting agreement (since 2014), the weighted system was replaced by a single per Member State. Below is a chronology of the voting system development.

Before 1 May 2004 there were only 15 Member States and the voting system was limited to 87 votes as follows:
Germany, France, Italy, UK 10 votes each
Spain 8 votes
Belgium, Netherlands, Portugal, Greece 5 votes each
Sweden, Austria 4 votes each
Finland, Denmark, Ireland 3 votes each
Luxembourg 2 votes

Enlargement in 2004 increased membership of the EU by a further ten countries. The weighting of votes inevitably had to change to accommodate the new members. There was a new system of weighting from 1 November 2004, with the total number of votes 321, broken down as follows:
Germany, France, Italy, UK 29 votes each
Spain, Poland 27 votes each
Netherlands 13 votes
Belgium, Czech Republic, Greece, Hungary, Portugal 12 votes each
Austria, Sweden 10 votes each
Denmark, Ireland, Lithuania, Slovakia, Finland 7 votes each
Cyprus, Estonia, Latvia, Luxembourg, Slovenia 4 votes each
Malta 3 votes

From 1 November 2004 a Qualified Majority Vote occurred in the following circumstances:

  • a majority of Member States (in some cases a two-thirds majority) approved the measure; and
  • a minimum of 232 votes was cast in favour of the measure – this represented 72.3% of the votes available (and is roughly the same as the previous 62 votes from 87).

In addition, under the new system Member States were able to ask for confirmation that the vote represented at least 62% of the total population of the EU. If not, then the measure would not be adopted.

From 1 January 2007, reflecting the membership of Romania and Bulgaria, the allocation of votes under the procedure was as follows:
Germany, France, Italy, UK 29 votes each
Spain, Poland 27 votes each
Romania 14 votes
Netherlands 13 votes
Belgium, Czech Republic, Greece, Hungary, Portugal 12 votes each
Austria, Sweden, Bulgaria 10 votes each
Denmark, Ireland, Lithuania, Slovakia, Finland, 7 votes each
Cyprus, Estonia, Latvia, Luxembourg, Slovenia 4 votes each
Malta 3 votes

Following the addition of Croatia as the newest Member State on 1 July 2013, the votes under the procedure were as follows:
Germany, France, Italy, UK 29 votes each
Spain, Poland 27 votes each
Romania 14 votes
Netherlands 13 votes
Belgium, Czech Republic, Greece, Hungary, Portugal 12 votes each
Austria, Sweden, Bulgaria 10 votes each
Denmark, Ireland, Lithuania, Slovakia, Finland, Croatia 7 votes each
Cyprus, Estonia, Latvia, Luxembourg, Slovenia 4 votes each
Malta 3 votes

Before the UK left the EU, this amounted to 352 votes and a qualified majority would be reached either:

  • where a majority of the 28 Member States votes in favour (sometimes even two-thirds); and
  • at least 260 of the 352 votes were in favour of the measure.

The definition of qualified majority is thus a double majority so that, to be adopted, an act must have the support of at least 55% of the EU Member States and must represent at least 65% of the population of the EU; and a blocking minority must include at least four Member States. In other words, whilst this meant a majority of 16 Member States out of 28 before Brexit, this means a majority of 15 out of 27.

However, it has been argued that whilst post-2014 the double-majority voting system empowered all Member States – big and small alike – the same cannot be said since Brexit. It appears that (based on the calculation of the Banzhaf indices), since the UK exited the EU, the strongest (i.e. more populated) countries appear to have grown stronger, gaining further voting power; whilst the smaller countries may have become weaker, in what is termed an unequal redistribution process of voting power.

2.2.2 The European Council

The European Council is not a creation of the Treaties. It was devised and agreed on by a meeting of heads of state in 1974 and was formally put in place in the Single European Act in 1986. Following the Lisbon Treaty, it is now an established institution of the EU governed by Articles 235–236 TFEU and Article 15 TEU.

Its role is to ‘provide the Union with the necessary impetus for development [and to] define the general political guidelines’.

It involves twice-yearly meetings of the heads of state of the Member States together with their Foreign Ministers. These are usually referred to as ‘summits’. These involve general policy-making and also the two pillars that do not have the force of Community law: common foreign and security policy, and police and judicial co-operation in criminal matters.

One other useful aspect of these meetings is that they provide the means of assuring approval of measures proposed by the Commission. If the European Council has accepted proposals then it is unlikely that they would be rejected by vote in the Council of Ministers. This helps to streamline the process of legislation.

Presidency of the European Council used to rotate with the presidency of Council between Member States. The Lisbon Treaty created a new system. The President is elected by a qualified majority of the members for a period of two-anda-half years. This term of office can be extended once. The President represents the EU in common foreign and security policies, and also acts as chair.

2.3 The European Commission

The Commission is governed by Articles 244–250 TFEU. Of all the EU institutions it probably has the clearest claim to being a supranational body of the type envisaged by the founders of the Treaties.

The Commission actually refers to two groups of people. It refers to the Commissioners themselves, those representatives appointed from Member States who run the various departments. It also of course refers to the whole staff of the Commission, administrative officials, translators, secretaries and other staff, amounting to over 32,000 with more than 65% working in Brussels.

The Commission essentially acts as an executive arm of the EU and is also known as the ‘Guardian of the Treaties’. It is sometimes compared to a civil service, although in fact it has much broader powers and roles than any conventional civil service.

2.3.1 Composition and appointment

Prior to enlargement the Commission comprised 20 Commissioners, as per former Article 213 ECT, and there was to be at least one from each Member State but not more than two. In practice, then, traditionally the larger states – France, Germany, Italy, the UK and Spain – had two Commissioners while the other states had one each. With the impending enlargement, the Treaty of Nice provided in an amended Article 213(1) that from 1 January 2005 the Commission should include one national of each Member State. This meant that the larger states would lose their extra Commissioner, which was to prevent the Commission from becoming too large and unmanageable with enlargement. The amended Article also stipulated that once there are 27 Member States there would be fewer Commissioners than Member States, so that the Council would introduce a rotation system. However, the Accession Treaty for the entry of Romania and Bulgaria provided for both to gain a Commissioner, and Croatia also gained a Commissioner on entry in 2013 so the number rose to 28 by 2013.

Under the Treaty of Lisbon, a.k.a. the Reform Treaty, the number of Commissioners is meant to be reduced to a number equivalent to two-thirds of the number of Member States, these being selected on a strict rotational basis (Article 17(5) TEU). However, to date, there are 27 commissioners, one for each Member State of the Union.

The Commission is appointed for a five-year fixed-term presidency (Article 17(3) TEU). Commissioners are chosen on grounds of general competence, from persons ‘whose independence is beyond doubt’. In practice they are nominated by their Member States, whose approval process and appointment subsequently involves the European Parliament.

Whilst they are nominated by their Member States, commission are, in carrying out their duties required by Article 245 TFEU to take an oath, ‘a solemn undertaking’ to respect the Treaties and the Charter of Fundamental Rights; to be independent; to neither to seek nor to take instructions from their Member States’ governments; and to refrain from any action incompatible with said duties. The Member States also undertake not to influence the Commissioners. Each Commissioner is appointed for a five-year term and is assigned a Directorate General with specific responsibilities, for example for employment or transport or the Internal Market.

2.3.2 The President

The Commission is headed by a President. The President is nominated by the governments of the Member States and must subsequently be approved as part of the whole Commission by the European Parliament.

Post-Treaty of Lisbon, for the first time in Union history, in 2014 the appointment of the Commission President became politicised and somewhat democratised as EU citizens’ vote for members of the European Parliament could influence the latter’s choice of Commission President. The main European Parliament political parties each put forward their designated candidate for nomination – this practice, which originates from Germany, is known as the Spitzenkandidaten process. The European People’s Party (EPP) lead candidate for the Commission 2014–2019 term won as Jean-Claude Juncker became President. However, it seems that the process, which was deemed to bring more democratic legitimacy to the system, was cast aside in the 2019 elections, in favour of a surprise nominee – and eventual winner Ursula von der Leyen – who was put forward by a Member States alliance, allegedly because some heads of national governments did not approve of the Spitzenkandidaten process.

The President has the authority to decide on the internal organisation of the Commission and the President clearly has the authority within this to allocate responsibilities among the Commissioners.

2.3.3 The role of the Commission

The Commission is collegiate in character and tends to act on the basis of simple majority votes.

Its key responsibilities are:

  • To initiate legislation. The Commission proposes legislation and can draft proposals on anything covered by the Treaties which it presents to Parliament and the Council. In doing so it takes advice from the Economic and Social Committee and the Committee of the Regions. So that it does not interfere in issues that can be better dealt with by the Member States themselves, the Commission also operates according to the ‘subsidiarity’ principle.
  • To enforce the law. The Commission has often been referred to as the ‘watch-dog’ or ‘guardian’ of the Treaties. Article 4(3) TEU demands that all Member States are obliged to achieve the objectives of the Treaties. The Commission must then ensure that all Member States are applying EU law properly. It can then deal with breaches of EU law by Member States through Article 258 proceedings (reasoned opinion for failure to fulfil an EU obligation). Initially the Commission would use the ‘infringement procedure’ but if a Member State fails to respond, then it can take the matter to the ECJ.
  • To implement policy and the EU budget. The Commission has executive functions. As a result, it is responsible for managing policy. For instance, traditionally it had a key role in EU competition law. It is also responsible for the compulsory budget. Even though it is actually national and local authorities that usually spend the money the Commission still has supervisory responsibility. In particular, the Commission would be responsible for the European Social Fund and the Regional Development Fund.
  • To represent the EU internationally. The Commission speaks for all the Member States in international meetings such as in the World Trade Organisation or the United Nations. It also negotiates international agreements for the EU, an example being the Cotonou Agreement which is a trade agreement between the EU and certain developing nations in Africa and the Caribbean.

Through a process known as comity the Council can delegate power to the Commission for it to produce detailed regulations following the passing of a framework regulation by the Council.

It must be remembered, though, that the Commission is also accountable to Parliament in certain ways. Parliament is able to pass a motion of censure on the Commission under Article 234 TFEU, causing the Commission to resign. This in fact happened in 1999 following a lengthy inquiry into fraud.

2.4 The European Parliament

In the original EC Treaty, Parliament was identified as the Assembly and its role was to ‘exercise the advisory and supervisory powers’ conferred upon it. It was not a democratically elected body, as it was made up of appointed nominees from the Member State governments. The limited powers of the Assembly were another reason for the complaint of a democratic deficit, since it had no legislative power and could only act in a consultative capacity.

Since 1979, however, the European Parliament has been directly elected by the citizens of the European Union to represent their interests. Every EU citizen can vote and stand as a candidate in the parliamentary elections. MEPs are elected for a period of five years. They are organised across seven loose political groupings, which at the time of printing, are EPP (European People’s Party); S&D (Progressive Alliance of Socialists & Democrats); Renew (Renew Europe); Greens/EFA (Greens/European Free Alliance); ID (Identity & Democracy): ECR (European Conservatives & Reformists); and the Left. Each grouping includes representatives from a number of Member States, rather than according to any national interests. In this respect there is no mandatory voting for Member State interests, as MEPs are representatives and not delegates. Voting is on a simple majority basis. There are also a number of specialist parliamentary committees. Parliament also elects its own President and various officials.

The European Parliament expresses the democratic will of the Union’s citizens and is the voice of almost 450 million people; it is governed by Article 14 TEU and Articles 123–234 TFEU.

2.4.1 Composition

Prior to enlargement the total membership of Parliament was 626 MEPs. Membership in fact depends on the size and importance of the particular Member State. The Treaty of Amsterdam amended Article 189 to provide a maximum number of 700 MEPs, although with the impending first waves of enlargement, Article 189 was subsequently amended in the Treaty of Nice to provide for a maximum number of 732 MEPs.

The number of seats in the European Parliament for the 2009–14 term was 736. The allocation of MEPs by country from the start of the parliamentary term from 2009–14 was as follows:

Member State Number of MEPs each
Germany 99
France, Italy, UK 72
Poland, Spain 50
Romania 33
Netherlands 25
Belgium, Greece, Portugal, Czech Republic, Hungary 22
Sweden 18
Austria, Bulgaria 17
Denmark, Finland, Slovakia 13
Ireland, Lithuania 12
Latvia 8
Slovenia 7
Cyprus, Estonia, Luxembourg 6
Malta 5

In light of the delay in ratifying the Treaty of Lisbon, the 2009 elections for Parliament took place under the cap created by the Treaty of Nice. An amendment to the Treaty in 2011 allowed for 18 further MEPs to be elected prior to 2014 and Croatia on entry was allowed 12 MEPs. As a result, the membership of Parliament went up to 766. However, since the 2014 European Parliament elections, the number of MEPs was capped at 751 (750 MEPs and 1 President). The allocation by country, before Brexit, was as follows:

Member State Number of MEPs each
Germany 96
France 74
Italy, UK 73
Spain 54
Poland 51
Romania 32
Netherlands 26
Belgium, Greece, Portugal, Czech Republic, Hungary 21
Sweden 20
Austria 18
Bulgaria 17
Denmark, Finland, Slovakia 13
Ireland, Lithuania, Croatia 11
Latvia 8
Slovenia 7
Cyprus, Estonia, Luxembourg, Malta 6

Since 1 February 2020, with the withdrawal of the UK from the Union, the number of Members of the European Parliament has decreased to currently 705. Out of 73 UK seats, 27 have been reallocated through degressive proportionality, that is seats are allocated according to the size of a Member State’s population (e.g. the greater the population of a Member State, the larger number of MEP seats it will receive).

Whilst Member State did not lose any seats, Germany did not gain any additional seats, and 46 seats have yet to be allocated. They are being retained for the eventual enlargement of the Union should one or more of the candidate countries join the EU in the near future.

Member State Re-allocated seats Total seats
Austria +1 19
Belgium = 21
Bulgaria = 17
Croatia +1 12
Cyprus = 6
Czech Republic = 21
Denmark +1 14
Estonia +1 7
Finland +1 14
Germany = 96
Greece = 21
France +5 79
Hungary = 21
Ireland +2 13
Italy +3 76
Latvia = 8
Lithuania = 11
Luxembourg = 6
Malta = 6
The Netherlands +3 29
Poland +1 52
Portugal = 21
Romania +1 33
Spain +5 59
Slovakia +1 14
Slovenia = 8
Sweden +1 21

2.4.2 Location

Through historic development since the creation of the Union, the European Parliament’s seats are spread across three Member States. As formally recognised in Protocol 6 on the location of the seats of the institutions (annexed to the Treaties), the official seat of the European Parliament is situated in Strasbourg (France), where 12 plenary sessions are held throughout the calendar year. In other words, Parliament sits for one week each month (except August), in a plenary session during which its formal decision-making occurs; although it can also sit at other times when certain items require discussion. In addition, the EP’s main headquarters, activities, committees and political meetings are held in Brussels (Belgium); whilst its General Secretariat and administrative departments are located in Luxembourg.

2.4.3 The role of Parliament

Parliament currently enjoys three main powers:

  • It has a role in legislation. In the Ordinary Legislative Procedure (OLP and previously known as the co-decision procedure), now the most common method for introducing legislation, Parliament has an important role and can make amendments to legislative proposals, as set out in Article 294 TFEU. In some cases, the EP can exercise a veto (under the special legislative procedure as set out by the Treaties, such as the consent procedure). Under the co-operation procedure, which was introduced in the SEA and is now another special legislative procedure (SLP) known as the consent procedure, the European Parliament also has a consultative role, where the Council is required to seek the EP’s opinion although it is not bound to follow it. Special legislative procedures are not specifically defined and their use is determined by certain provisions of the Treaties. Nevertheless, their legal base is found in Article 289(2) TFEU. Parliament is further able to examine the annual work programme of the Commission.
  • It has a supervisory role over other EU institutions. This is particularly the case with the Commission. Parliament must approve each new Commission. It can also pass a motion of censure on the Commission. The effect of such a censure was seen in 1999 when it led to the resignation of the entire Commission. It is also able to send questions to the Council and also can express its views to each meeting of the European Council. The supervisory power of Parliament over the Commission has also been extended. Article 17(7) TEU requires that the commissioners, including the President and the High Representative of the Union for Foreign Affairs and Security Policy, are subject to the consent of Parliament subject to taking up office.
  • It has powers over the budget. Parliament is required to approve each annual budget. In the case of a failure to accept the budget the effect is dramatic since this includes the day-to-day payment of all officials in the institutions. The Treaty of Lisbon extended the power of Parliament over the budget. Now under Article 314(4) TFEU it has the power to amend any part of the budget, whether the compulsory or the non-compulsory.

Despite the increasing powers afforded by the Treaty of Lisbon, criticisms of democratic deficit keep being levelled against the European Parliament. There is still some discontent that the EP is unable to initiate legislation itself, despite the introduction of Article 225 in the TFEU which gives the EP a right of initiative – indeed, the provision enables the EP to request a legislative proposal from the Commission, however, the latter can decline provided it can justify its refusal to oblige. Also, the censure facility only covers the whole Commission and some feel there should be the power to censure and remove individual Commissioners. It is also felt by some that the Council should be more accountable to Parliament.

2.5 Court of Justice of the European Union (and the General Court)

2.5.1 The Court of Justice

The Court of Justice is not like any court in the English legal system. In fact, it is a court with few meaningful comparisons. As can be seen with the doctrines of primacy of EU law and direct effect for instance, the Court has played an absolutely vital role in the development of EU law.

The rules governing the Court of Justice are found in Articles 251–281 TFEU and by the Statute of the Court of Justice, as well as in the Protocol on the Statute of the Court of Justice of the European Community. The latter is added to the Treaty as an annex, and contains all the procedural rules of the Court. Despite the Court and the General Court (formerly known as the CFI – Court of First Instance) made representations at the Nice summit that they should be able to determine and adapt the procedure themselves, this was accepted. Therefore, the procedures are voted on by the Council, although since the Treaty of Nice, this is done with a Qualified Majority Vote rather than requiring a unanimous vote.

The composition of the Court

The Court of Justice of the European Union (CJEU) was formerly known as the European Court of Justice (ECJ) and still is often referred as such. The CJEU/ECJ is composed of juges rapporteurs (the judges) who are assisted by Advocates-General. An A-G assists judges in the performance of their functions, by providing reasoned, independent and impartial opinions on a case.

Article 253 identifies the requirements for appointment of both:

As a result of this the judges are chosen from high-level judges from the Member States or highly competent lawyers who are independent beyond doubt, so that they can be relied upon to show impartiality. Indeed, they must all swear an oath of impartiality.

In terms of numbers, before enlargement Article 19 TEU provided that ‘The CJEU shall consist of 15 judges’. This did in fact reflect the number of Member States and a judge for each one. The Treaty of Nice amends Article 221 to provide that the Court will consist of ‘one judge per Member State’, therefore today there are 27 judges for each of the Member States.

Both the judges of the CJEU and the Advocates-General are appointed by joint agreement of the governments of the Member States. Article 19 TEU stipulates that there shall be one judge appointed from each Member State. The Treaty also provides for the appointment of eight Advocates-General to assist the Court. Both judges and Advocates-General serve for a six-year term. There is a staggered re-appointment system and it is possible to be re-appointed for a further one or two periods of three years. Removal of a judge is possible only if all colleagues agree that the judge in question is unfit to serve.

Under Article 254 TFEU, the judges appoint a President from among themselves. The President serves for a three-year period. The President generally directs the business of the Court, appoints a specific juge rapporteur to manage a specific case, and tends also to deal with all interlocutory matters.

The role of the Court

By Article 13(2) TEU the Court of Justice and the General Court can only act in areas where jurisdiction has been specifically given to them in the Treaties.

There are three central objectives in the work of the Court:

  • to ensure that in application and interpretation the law is observed;
  • to provide a forum for resolving disputes between institutions, Member States and individuals;
  • to protect individual rights.

The Court hears five main types of action:

  • Article 267 references from Member States for a preliminary ruling on an interpretation of EU law (known also as indirect actions);
  • Article 258 actions against Member States for failing to implement Treaty obligations (a direct action known also as infringement proceedings);
  • Article 263 actions against an institution for abuse of power;
  • Article 265 actions against an institution for a failure to act;
  • Article 340 actions for damages against an institution that has been responsible for loss to the individual (for example where the Commission has failed to address a decision to a body engaging in anti-competitive practices and an individual suffers loss as a result).

Procedure

In the past, before the enlargement of the Union, most issues involving Member States or an institution were heard by a full court. There were some straightforward issues where it was possible for a bench of three or five judges to reach a decision. This was appropriate where the Court was limited in size to 15 judges. Obviously, enlargement means that a plenary session of the Court could involve a very large number of judges. Therefore, for the sake of efficiency, the Treaty of Nice allows the Court to sit as a ‘Grand Chamber’ of fewer judges instead of always having to meet in plenary session. In any case, Article 17 of the Statute of the Court of Justice provides that a decision of the Court is valid only if the Court comprises an uneven number of judges.

Decisions of the Court are thus based on a majority. Procedure is essentially inquisitorial. There is no provision for individual judges to deliver dissenting judgments.

Cases are submitted through the ‘registry’ and the President of the Court assigns the case to a specific judge to manage. In the first stage all parties make written submissions on which the judge writes a report and then passes everything on to an Advocate-General assigned to the case. The Advocate-General then produces a reasoned opinion for the Court. This does not have to be followed by the Court, though it may be. Following the introduction of the reasoned opinion, the judge prepares a draft ruling which is passed to the other members of the Court.

A public hearing of the action is held before the whole court in plenary session (or post-enlargement a grand chamber of the Court) or, depending on the type and complexity of the issue, a chamber of three or five judges. All parties can put forth their case and the Court can ask questions.

2.5.2 The General Court

Because of the excess workload of the Court of Justice and the long delays that resulted from this, a Court of First Instance (CFI) was created in the SEA. Since the Treaty of Lisbon this is now the General Court. The Court is governed by Article 256 TFEU which also identifies the types of case that could be heard by the Court.

The jurisdiction of the General Court when it was the CFI was limited to actions including staff cases and some actions under competition law.

An amendment in the TEU added another category: actions by natural or legal persons under either Article 230 or Article 232 (now Articles 263 and 265 TFEU), including anti-dumping cases.

Originally the Court was specifically excluded from hearing Article 267 references. Following the Treaty of Nice, the Court can hear references for preliminary rulings on specific areas although there are as yet no specific areas identified in the Treaty. The Treaty of Nice expanded the jurisdiction of the CFI, now the General Court, so that it can hear all actions under Articles 263, 265, 268, 270 and 272 TFEU that are not already attached to a ‘judicial panel’ or those that are required by the Statute of the Court of Justice to be heard in the CJEU. Judicial panels annexed to the General Court (formerly the CCFI) were another creation of the Treaty of Nice. The Council may set up such panels to hear specific types of action in specific areas to deal with issues speedily.

The regulations and requirements for the General Court are similar to those for the CJEU. Membership of the Court again is based on a representative from each Member State, with similar qualifications needed (although the General Court does not have Advocates-General). While independence again is an absolute requirement, members of the Court must merely ‘possess the ability required for appointment to judicial office’. From 2019, the membership of the General Court will increase from 28 to 56 judges.

In diagram form, the courts can be represented as shown in Figure 2.1.

A flow diagram depicts the work of the Court of Justice of the European Union and the General Court.

Long Description for Figure 2.1

The flow starts with member state governments appointing judges for a six-year period. It leads to the European Court of Justice hearing different types of actions. It leads to actions against member states, European Union institutions, and references from member states. In the General Court, one judge from each member state hears different actions, such as direct actions by natural and legal persons, and now jurisdiction is expanded after the Treaty of Nice.

Figure 2.1 The work of the CJEU and the General Court

2.6 The other major institutions

2.6.1 The Committee of Permanent Representatives (COREPER)

COREPER is the Committee of Permanent Representatives from all of the Member States. The name is an acronym from the French way of referring to the committee. The committee was not a product of the original Treaties. It was created by Article 4 of the Merger Treaty in 1965 and is now under Article 240 TFEU, and forms part of the decision-making procedure.

COREPER is tasked with preparing the work of the Council, which was deemed a necessity for consistency’s sake, given the fluid membership of Council, as a means of informing Ministers from the Member States and streamlining the process of legislation.

In this way, individual representatives prepare items of discussion at Council meetings and examine the Commission’s legislative proposals for the individual Ministers. It is therefore an integral part of the decision-making process, and generally, if a proposal can be agreed upon by COREPER before the Council meeting, it will be accepted by the latter without need for lengthy discussion.

2.6.2 The Court of Auditors

This body is not a product of the original Treaties. It was created by a Secondary Budget Treaty in 1975 inserted into the Treaty, and was eventually made a full institution in the TEU.

The Court of Auditors is an independent external body, whose basic role is to control and supervise the EU budget. It examines the accounts of all revenue and expenditure and, in effect, checks that the EU budget is correctly implemented by those institutions that are responsible for it. In this way, it can investigate the paperwork of any body handling EU funds and can carry out spot checks if appropriate. It also prepares an annual report for Parliament and the Council. It will in any case produce an ‘opinion’ before any financial measure is adopted. However, it does not have any legal powers of its own but passes information on to the other bodies for them to deal with.

The Court’s members are qualified professional auditors from each Member State; appointed by the Council. They are independent and chosen for these qualities, as they are tasked with looking after the interests of the EU taxpayers.

2.6.3 The European Economic and Social Committee (EESC)

The European Economic and Social Committee was first set up under the EC Treaty in 1957 as an advisory body to the institutions of the Union, and its role has been reinforced through the various amendment Treaties.

The EESC is governed by Article 13 TEU and 300 TFEU, which together stipulate it is to assist the political institutions of the Union, i.e. Parliament, Council and Commission. The Committee is composed of representative from socio-economic, civic, professional and cultural organisations, including employers, employed and civic society organisations. Membership is based on representation of various social or economic activities, and includes various interest groups such as farmers, carriers, dealers and craftsmen of different types.

The Committee’s purpose is to give advice, in the form of an opinion, to the Council, Parliament and Commission on social and economic matters. Since it is not an institution of the EU but a specialised body, there is no overall obligation to consult it. Nevertheless, under Article 304 TFEU, the EESC can be consulted by the EU political institutions whenever there is a specific obligation under the Treaties to do so. Consequently, where the EU institutions fail to consult on a matter where the Committee’s opinion is required, it is possible for the ECJ to annul the measure in question. Therefore, in practice, the Committee does have some degree of influence, and can in any case deliver ‘own-initiative opinions’, i.e. issue an opinion even where it has not been called for. Article 300 TFEU further adds that the EESC is not bound by any mandatory instructions, and its members are “completely independent in the performance of their duties, in the Union’s general interest”.

Upon proposals from the Member States, Council appoints the members of the committee for a period of five years (renewable). Representation on the Committee is based on the size of the Member State. Article 301 TFEU identifies that (after enlargement) the Committee shall not exceed 350 in number, and the allocation of EESC representatives by country is as follows:

Member State Members
Germany, France, Italy 24 each
Poland, Spain 21 each
Romania 15
Austria, Belgium, Bulgaria, Czech Republic, Greece, Hungary, The Netherlands, Portugal, Sweden 12 each
Croatia, Denmark, Finland, Ireland, Lithuania, Slovakia 9 each
Estonia, Latvia, Slovenia 7 each
Cyprus, Luxembourg 6 each
Malta 5

2.6.4 The Committee of the Regions

Although the Committee of the Regions was not a product of the original Treaties, it is a recognised specialised body of the European Union and is set to assist the institutions of the EU (Article 300 TFEU). The COR was created in 1994, after the Maastricht Treaty, as an advisory body representing local and regional interests, and since the Treaty of Lisbon is governed under Articles 305–307 TFEU.

The COR will be consulted within the Union legislative process on matters of regional and local impact, such as education, public health, culture and other matters of social important. This provides European regions and cities a voice and a position in the law-making process, ensuring that the European Commission, Council and Parliament consider local and regional government concern. Failure to consult with COR, e.g. to give it a chance to review and give an opinion on a legislative proposal of relevance to its interest, could lead to COR taking action against the EU institutions before the Court of Justice of the EU. It is also possible for the COR to initiate their own opinion.

By Article 300 TFEU, its membership is composed of elected and/or politically accountable members of local and regional bodies (although not national government); e.g. mayors, elected representatives, and leaders of regional authorities and cities across the 27 Member States. COR members, upon proposals from EU Member States, are appointed by the Council for a renewable 5-year period; and the ceiling on its overall numbers is at 329 since the departure of the UK from the EU (Council Decision (EU) 2019/852), allocated on a proportionate basis, i.e. according to the population size of each Member State – the same numbers as for the EESC.

2.6.5 The European Central Bank

The European Central Bank (ECB) was a creation of the TEU in which provision was made in the Protocol (and now under Articles 282–284 TFEU) to have a Central Bank to act for the EU.

The purpose of setting up such a body is clear. It is an essential element of the policy of European Monetary Union (EMU) and the move to a single currency.

The bank has been in place since 1 January 1999 and has responsibility for monetary policy in the EU. Its primary aim is to maintain price stability. The ECB is an independent body that has legal personality and sole competence for issuing euro currency.

Under Article 284(5) the ECB has an automatic right to be consulted on legislative and regulatory proposals, at Union and national level, in areas that pertain directly to the ECB’s responsibilities.

2.6.6 The European Investment Bank

The European Investment Bank, governed by Articles 308–309 TFEU and Protocol 5 on the statute of the European Investment Bank (annexed to the Treaties), is a multilateral financial institution and ‘the lending arm of the European Union’. The EIB has legal personality and its membership is composed of the 27 Member States of the Union.

The EIB is tasked with promoting the ‘balanced and steady development of the internal market’ in accordance with the objectives and interests of the Union. Its basic mission is to invest in projects whose funding/financing is, by their size or nature, beyond the means of any single Member State. Therefore, as per Article 309 TFEU, the EIB shall operate on a non–profit-making basis across all sectors of the economy, and grant loans or provide guarantees for the financing of:

  • projects for developing less-developed regions;
  • projects for modernising, converting or developing activities for the good functioning and development of the internal market;
  • projects of common interest to several Member States.

The EIB is not financed by the EU budget but by borrowing in the financial markets and also from the Member States. It invests in projects according to strict criteria, whereby the project must help achieve EU objectives (e.g. making small businesses more competitive); help mainly disadvantaged regions; or contribute to attracting other sources of funding.

2.6.7 The European Ombudsman

The office of the European Ombudsman was established by the Maastricht Treaty (1992) and is governed by Article 228 TFEU and Regulation (EU, Euratom) 2021/1163 laying down the regulations and general conditions governing the performance of the Ombudsman’s duties (Statute of the Ombudsman’s duties).

The European Ombudsman is elected by the European Parliament for a period of five years (renewable).

They are an intermediary between European citizens and EU institutions, who act independently and listen to complaints from EU citizens, as well as investigate examples of maladministration. EU citizens have a two-fold fundamental right:

  • to apply to the EO under Article 20(2)(d) TFEU);
  • to good administration under Article 41 of the Charter of Fundamental Rights.

EU citizens can therefore bring a complaint to the EO against EU institutions, bodies, agencies, offices and representatives, whose action are deemed a maladministration. In an EU context ‘maladministration’ can concern:

  • unfairness;
  • discrimination;
  • abuse of power;
  • lack of information or refusal to give information;
  • unnecessary delay in making decisions;
  • using incorrect procedures.

The Ombudsman can refer matters to the other institutions to take appropriate action but will not investigate a complaint that has been the subject of a court case.

SUMMARY

  • There are four main EU institutions.
  • Council and Parliament are the two legislators – upon proposals from the Commission.
  • The Commission prepares draft legislation, is the major administrator and is also ‘the watchdog of the Treaties’.
  • Parliament – is the only directly elected body. IT has law-making powers as well as some control over the budget.
  • The Court of Justice, which provides preliminary rulings on interpretation, hears actions against Member States for breaches of EU law, hears judicial review for abuses by the institutions and for their failure to act, and actions for damages by citizens.
  • There are also other institutions including COREPER, the Economic and Social Committee, the Committee of the Regions, the Court of Auditors and the European Central Bank – and the Court of Justice is also supported by the General Court.

Further reading

Articles

  • Bradley, K, ‘Institutional Design in the Treaty of Nice’ (2001) 38 CMLR 1095.
  • Gábor, J, ‘Impact of Brexit on Voting Power in Council of the European Union’ (2020) 3 Open Political Sci 192–197.
  • Commentaries
  • Gray, A, Barigazzi, J and de la Baume, M, ‘Who Killed the Spitzenkandidat? No Shortage of Suspects in Drama Over How EU Should be Governed’ Politico, 5 July 2019.
  • Tilindyte, L, ‘Election of the President of the Commission. Understanding the Spitzenkandidaten Process’ Briefing Paper. EPRS European Parliamentary Research Service, April 2019.

Books

  • Craig, P and de Burca, G, EU Law: Text, Cases and Materials (OUP, 2020).
  • Douglas-Scott, S, Constitutional Law of the European Union (Longman, 2002), chaps. 2 and 5.
  • Schütze, R, European Union Law (OUP 2021).
  • Ward, I, A Critical Introduction to European Law (3rd edn, CUP, 2011), chap. 1.

3 The sources of EU law

DOI: 10.4324/9781003218562-3

3.1 Primary sources – the Treaties

3.1.1 The importance of the Treaties

The Treaties are the most significant source of EU law and primary legislation, which is the equivalent to a constitution and why they are also referred to as ‘the founding Treaties’. The original founding Treaties – the ECSC, the EURATOM Treaty and the EC Treaty (now TFEU) – are all primary law which have been amended and now been consolidated by the Treaty of Lisbon in the form of the Treaty on the European Union (TEU) and the Treaty on the Functioning of the European Union (TFEU).

As the Community expanded and the Union was created, a number of related Treaties have been introduced. These all have the force of Community [now EU] law and create enforceable rights and obligations. They include:

  • the various Accession Treaties expanding the original Community territorially;
  • the Merger Treaties 1965;
  • the Single European Act 1986 (which put in place the processes to eventually achieve the Common Market);
  • the Treaty on European Union 1992 (which created the Union and its three-pillar structure);
  • the Amsterdam Treaty 1997 (which rationalised the existing structure and renumbered the original TFEU);
  • the Treaty of Nice 2000 (which focused on institutional reform);
  • the Charter of Fundamental Rights of the European Union 2001 (which provides a codified catalogue of socio-economic, political, civil and human rights for the protection of individuals across the EU Member States);
  • the Treaty of Lisbon, which was introduced in place of the rejected Constitutional Treaty (has introduced further institutional reform as well as amending both the TEU and the EC Treaty and renaming the latter as the TFEU. It also gave legal binding force to the CFR).

Whilst the TEU sets out the objectives, governing principles and institutional framework of the European edifice, the TFEU, as the title suggests, provides the structural foundations for the more detailed organisation and function of the EU.

The TEU and TFEU, along with the CFR, form the constitutional basis of the EU legal order. In addition, with general principles of EU law and international agreements, the Treaties form the primary legislation block of the EU. As such, all subsequent law must fulfil the objectives of those founding Treaties, respect the provisions of the CFR and the general principles of EU law (as developed by the ECJ), as well as the terms of the international agreements that form part of the EU legal order.

The Treaties are directly applicable as they become part of the Member States domestic legal system without the need for national implementing measures.

3.2 Secondary sources – legislation under Article 288 TFEU

3.2.1 Introduction

‘Secondary legislation’ is a collective term that is used to describe all of the various types of law that the institutions can make and the legal instruments which they can use to regulate the policy areas within the competence of the EU. Secondary legislation is of major importance since it is the way in which EU law is expanded and developed out of the broad principles contained in the Treaties themselves. Nevertheless, it is also important to remember that secondary legislation is still subordinate to the primary law of the Treaties and must only be used for the furtherance of the objectives of the Treaties. As a result, the legislation cannot amend, repeal or alter the scope of the Treaties.

In this way the institutions may only act in secondary legislation:

  • in order to carry out the tasks assigned to them by the Treaties;
  • in strict accordance with the provisions of the Treaties for the fulfilment of the objectives of the Treaties;
  • and only within the strict limits of the powers that are actually conferred upon them in the Treaties, and specifically those identified in Article 288 of the TFEU.

The introduction of the ordinary legislative procedure (OLP) under Article 289 TFEU has created a distinction between a legislative act (thus adopted by OLP with the participation of the EP, Council and Commission) and a regulatory act (which usually originate from the Commission through delegated powers). However, it is Article 288 TFEU that defines the role of the institutions in producing legislation:

As well as defining the power of the institutions to introduce legislation, Art 288 also defines the different kinds of legal instruments or legal acts in which secondary legislation can take form. As can be seen from the wording of Art 288, it is their scope and effect which distinguish them from each other.

3.2.2 The different types of secondary legislation

Binding secondary legislation

Regulations

Regulations are defined in paragraph 2 of Article 288:

The terminology used in the Article needs to be understood in order to appreciate the scope and effect of a Regulation:

  • ‘General application’ (otherwise referred to as ‘general applicability’) simply means that the measure applies generally to all Member States.
  • ‘Binding in its entirety’ means that the Member States have no choice whether to give effect to the measure. They are bound by the regulation in its entirety.
  • ‘Directly applicable’ (again commonly referred to as ‘direct applicability’) means that the measure automatically becomes law in each Member State on the date specified. The consequence of this is that there is no requirement for the state to implement the measure (see section 8.1).

It would be easy from the description given to compare regulations with Acts of Parliament in the UK. They automatically become law on the date specified and are absolutely binding.

Obviously, they also operate in this sense slightly differently from the Treaties themselves. Having a dualist constitution, the UK only became bound by the Treaties once they had been ratified and incorporated into UK law by the European Communities Act 1972. Once the UK had signed the Treaties and incorporated them into English law there is no similar requirement for the introduction of Regulations into English law. They are binding once introduced.

Regulations are also capable of creating rights and obligations which are then directly enforceable in the national courts through the principle of direct effect (see Chapter 8): Leonesio v Ministero dell’Agricoltora & delle Foreste (Case 93/71) [1972] ECR 287.

Directives

Directives are defined in paragraph 3 of Article 288:

Again, the wording in Article 288 indicates the scope and effect of Directives, although in this case there are more significant problems in terms of their possible effects as legislative measures.

The two key aspects to the paragraph are:

  • ‘binding as to the object to be achieved’; and
  • ‘shall leave to the national authorities the choice of form and methods’.

The wording here is significant. It indicates that Directives are quite unlike Regulations which are directly applicable and demand absolute uniformity. Instead, Directives are not directly applicable but are used to ensure that Member States adapt their own laws for the application of common standards. They are about the harmonisation of Member State law on specific issues.

As such, they leave an element of discretion to the Member States and allow the Member States to select what is for them the most appropriate method of implementation. However, they are bound to do so within a set deadline.

Because they are harmonising measures they are mainly used in those areas where the diversity of national laws could prevent the proper establishment or even the effective functioning of the Single Market. A classic example of this harmonising process can be found in the so-called sectoral Directives introduced for the recognition of different professional and vocational qualifications under Article 49 for the furtherance of freedom of establishment. This also applies to the more generalised Directives 89/48 and 92/51 and the ‘Slim Directive’ 2001/19 (see Chapter 13).

In contrast, then, whereas a Regulation is applicable to all Member States as well as individual citizens alike, a Directive is really only intended to create legal obligations on the Member States. In this way directives were not originally seen as being intended to create rights that could be directly enforced by individuals. Nevertheless, to avoid the possibility of EU law being ignored by the Member States the ECJ has created the means to ensure that they can be enforced. This has been controversial and a more detailed explanation is given in Chapter 8, but the main ways are:

  • Vertical direct effect. This is a process by which individuals may enforce an unimplemented Directive, provided that the date for implementation has passed (Pubblico Ministero v Ratti (Case 148/78) [1979] ECR 1629) and the claim is against either the state (Marshall v Southampton and South West AHA (No 1) (Case 152/84) [1986] QB 401) or an ‘emanation of the state’ (Foster v British Gas plc (Case C–188/89) [1991] 1 QB 405).
  • Indirect effect. This is the principle developed in Von Colson v Land Nordrhein-Westfalen (Case 14/83) [1984] ECR 1891 and Marleasing SA v La Commercial Inter-nacional de Alimentacion SA (Case C–106/89) [1990] ECR I–4135. The ECJ has held that, because Article 10 (now Article 4(3) TFEU) of the EC Treaty demands that Member States fulfil all Treaty obligations, national courts should interpret national law so as to give effect to the Directive whether it is ineffectively implemented or not implemented at all.
  • State liability. This principle, stemming from the case of Francovich v Italy (Cases C–6 and 9/90) [1991] ECR I–5357, holds that while there can be no horizontal direct effect based on a Directive as between ordinary individuals, an individual who has suffered loss as a result of the Member State’s failure to implement a Directive may claim damages from the state.
Decisions

Decisions are defined in paragraph 4 of Article 288:

The two key elements of the definition are:

  • ‘binding in its entirety’; and
  • ‘those to whom it is addressed’.

In terms of scope and effect, obviously the first point about a decision is its effect. A decision is immediately and totally binding on the party to whom it is addressed. As a result of this it is equally clear that a decision is capable of creating obligations that are then enforceable by third parties. For instance, see Grad v Finanzamt Traustein (Case 9/70) [1970] ECR 825.

The next point to make is that a decision is clearly not generally applicable as it may be addressed to a limited range of parties and not to the EU generally.

What is also clear is that decisions are the least easy form of legislation to define. They could be legally binding measures created according to a specific legal form. However, they could also be non-binding, informal acts which lay down guidelines. A common context for the use of decisions has been in EU competition law. (See Chapter 16.)

Non-legally binding secondary legislation: recommendations and opinions

Article 288 also gives the Commission the power to ‘formulate recommendations’ and also to ‘deliver opinions’. The Article also identifies these as having no binding force.

As law, then, such measures can be seen as ineffective. Nevertheless, they are a useful means of clarifying issues in a less formal way than by introducing binding legislation.

While the measures are not enforceable as law, it is possible for them to have a persuasive effect on the Court of Justice in its decision-making. In fact, in Grimaldi v Fonds des Maladies Professionelles (Case 322/88) [1989] ECR 4407 the ECJ considered that national courts were bound to take recommendations and opinions into account in deciding cases. However, it is unlikely that this would be followed in practice.

General applicability Direct applicability Direct effect (see Chapter 8)
Treaty Articles These apply generally throughout the whole EU (so are generally applicable) Once a Treaty is incorporated there is no need for further enactment of Articles Will have if they conform to the Van Gend (1963) criteria
Regulations These apply generally throughout the whole EU (so are generally applicable) These require no further implementation (so are directly applicable) Will have if they conform to the Van Gend (1963) criteria
Directives Usually addressed to all Member States (in which case are generally applicable) These are an order for Member States to comply (so need implementation and are not directly applicable) Vertical direct effect only (if unimplemented/ incorrectly implemented and past implementation date)
Decisions Addressed to particular individuals (so are not generally applicable) These are an order that must be complied with by the addressee May confer rights on other individuals affected by them - so can be directly effective

3.3 Tertiary sources

3.3.1 The case law of the CJEU (formerly the ECJ)

The CJEU (formerly ECJ) has played a vital role in the development of EU law. The Article 267 procedure is the major means by which the application of EU law in the Member States is tested (see Chapter 6). References from national courts under this procedure lead to binding interpretations of Treaty provisions and legislation.

The importance of the Court in illuminating principles of EU law is obvious since the Treaties are framed in broad terms and cover broad principles. To a degree, the same point can be made of the secondary legislation. The CJEU adds detail and context to these broad principles and provides more precise principles for the national courts in the Member States to follow.

The CJEU is unlike any court that we are familiar with in the UK. In character it is based on the continental ‘civil’ or ‘Roman’ law systems. As such, there is no strict system of binding precedent as exists in English law and, in theory, the Court is not bound by its past decisions, as an English court would be.

In this way the Court could be said in the strictest sense to have moral rather than legal authority, and in a technical sense the Court’s decisions could be argued not to be a formal source of law. Nevertheless, a number of points could be made:

  • first, it is true that the Court will not depart from its past decisions without good reason;
  • second, in its reasoning and in its judgments the courts as well as the General Court have shown a remarkable consistency over the years;
  • in any case, the rules in CILFIT v Ministry of Health (Case 283/81) [1982] ECR 3415 on application of the preliminary reference procedure under Article 267 in essence prevent repetitious references by Member States trying to gain different rulings on the same principle of law;
  • finally, the Court in any case has proved to be very ‘legislatively active’ in its eagerness to achieve the ‘effet utile’ (effectiveness) in ensuring the attainment of the objectives of the Treaties.

In this sense the case law of the CJEU is in fact a major source of EU law and has been a key element in the development of EU law in two ways:

  • It has defined the principles that apply in all of the main areas of substantive law, e.g. the ‘Four Freedoms’, discrimination law, competition law, etc. (see Chapters 10–18).
  • It has ensured that the objectives of the Treaties are achieved in the Member States by developing the principles of supremacy and direct effect (see Chapters 7 and 8).

3.3.2 General principles of EU law (GPEUL)

There was nothing in the original Treaties that directed the ECJ (now the CJEU) to apply general principles of law in deciding cases. Article 6(1) of the TEU, as amended by the Treaty of Amsterdam, does identify that the Union is founded on principles of liberty, democracy and the rule of law as well as respect for human rights and fundamental freedoms, and also principles that are common to the Member States.

However, the Court has developed a number of unwritten principles that it will use when it interprets the Treaties and the secondary legislation. In doing so the Court relied on the authority of Article 220 which obliges it to interpret provisions so as to ensure that the law is observed by the Member States. In this way the general principles of law have been recognised as binding on the institutions, the Member States and, indeed, on individual citizens.

Subjecting interpretation of the law to general principles is not a novel idea. In fact, the practice is a familiar one in those states that have a ‘civil’ or ‘Roman’ law tradition. The general principles of law are in essence a statement of essential values and basic standards which are broad enough to be generally acceptable as principle. The process itself is not particularly controversial. It is the application of the general principles in specific situations that has been felt to be so at times.

Because a lot of EU law is essentially administrative, certain of the principles have derived quite naturally from the administrative law of both France and Germany. Nevertheless, some of the principles have their origins in UK law.

The main ones are:

  • proportionality;
  • equality;
  • legal certainty;
  • natural justice;
  • the protection of fundamental human rights;
  • subsidiarity.

Proportionality

Proportionality is a concept that comes from German administrative law and is known as ‘verhaltnismassigkeit’. The basic principle is that any measure or any action taken must be proportionate to the actual end to be achieved. A simpler explanation would be to say that nothing should be done that is more than is necessary to achieve the end.

The idea of applying the principle of proportionality in EC (now EU) law first came about in the Internationale Handelsgesellschaft case (Internationale Handelsgesellschaft GmbH v Einfuhr und Vorratsstelle fur Getreide und Futtermittel (Case 11/70) [1970] ECR 1125). Here, the ECJ adopted the principle in the following terms:

The principle of proportionality is now laid down in Article 5 TEU. The Court will apply the principle in relation to legislation, for instance by determining whether the legislation goes beyond what is necessary to achieve the actual purpose in the Treaty provision behind the legislation.

The Court also applies the principle when reviewing acts of the institutions, again, for example, in determining whether the action imposes too great a burden for the actual breach of EU law. An obvious context for this is the fines imposed for breaches of Articles 101 and 102 (see Chapter 16).

Another way in which the Court has exercised the principle is in reviewing the actions of Member States when claiming derogations under the various freedoms, particularly those in Article 36 in relation to the free movement of goods (see Chapter 14) and under Article 45(3) and Directive 2004/38 in relation to the free movement of workers (see Chapter 12).

Equality

The concept of equal treatment or non-discrimination is not just a general principle; it is also one of the founding principles of the EU. The TFEU includes three specific prohibitions against discrimination:

  • Article 18 prohibits any discrimination based on nationality (this is a base Article that also operates behind the various Treaty Articles creating the Common Market through the ‘Four Freedoms’).
  • Article 157 demands that men and women shall receive equal pay for equal work (and has subsequently been extended to cover all discrimination between the sexes as well as other areas such as race and religion).
  • Article 40 prohibits discrimination between producers and consumers in relation to the Common Agricultural Policy (CAP).

Besides this, the principle of equality was extended in the Treaty of Amsterdam, so the Treaty now includes a general aim of ‘equality between men and women’. This is a major development since it is not restricted to work as it previously was. An even more impressive development gives the Council the power to legislate on discrimination in a much more general sense. Article 19 TFEU now allows the Council the power to take action to remove discrimination based on ‘sex, race or ethnic origin, religion and belief, disability, age, and sexual orientation’. In fact, even before this the Commission was active in tackling discrimination and promoting equality. While English law included no specific provisions for tackling sexual harassment (so that women claiming had to use the residual category of ‘subjecting to any other detriment’ under s6(2)(b) of the Sex Discrimination Act 1975) a Commission Code of Practice had defined ‘sexual harassment’ as any ‘unwanted conduct of a sexual nature, or other conduct based on sex affecting the dignity of women and men at work’.

The ECJ has also been proactive in combating discrimination and advancing equality. In implementing the principle in Article 157 the Court has identified in Bilka-Kaufhaus GmbH v Weber von Hartz (Case 170/84) [1986] ECR 1607 that unequal pay can only be accepted if it is based on objective justification. In defining ‘objective justification’ the Court also relied on the principle of proportionality (see above). The idea of objective justification itself has subsequently been extended to apply to any inequality in Graff v Hauptzollamt Kohn-Rheinau (Case C–351/92) [1994] ECR I–3361.

Similarly, while English courts have accepted the legitimacy of discrimination against both transsexuals and gay people, the ECJ has been more prepared to apply the principle of equality in such cases. In P v S and Cornwall County Council (Case C–13/94) [1996] All ER (EC) 397 the Court applied the principle of equality to the dismissal of a transsexual. While in Grant v South West Trains Ltd (Case C–249/96) [1998] All ER (EC) 193 the Court did not feel bound to apply the principle to same-sex couples, who it felt were not in an ‘equal situation’ to heterosexual couples; this now falls under Directive 2000/78.

The Court has in any case already taken the principle of equality to extend to discrimination on religious grounds in Prais v The Council (Case 130/75) [1976] ECR 1589 (see further discussions in Chapters 17 and 18).

Legal certainty

This is not a novel concept and it is one that is familiar to most legal systems. The basic principle is that the law in its application must be both certain and predictable. This was identified at a very early stage in Da Costa en Schaake NV v Nederlandse Belastingadministratie (Cases 28 to 30/62) [1963] ECR 61. The Court has subsequently stated in Officier van Justitie v Kolpinghuis Nijmegen BV (Case 80/86) [1987] ECR 3969 that it is the duty of national courts to interpret EU law in such a manner that is ‘limited by the general principles of law’ and also that in particular national courts should observe ‘the principles of legal certainty and nonretroactivity’.

There are a number of potential consequences of applying the principle. One obvious consequence is that there should be no retroactive laws. This indeed was at least partly the case for refusing to say that Article 157 was retrospectively directly effective in Defrenne v SABENA (No 2) (Case 43/75) [1981] All ER 122. Of course, the judgment was affected by the objections by both the UK and Irish governments and the ECJ did accept that this principle would only apply in exceptional cases where extreme difficulties would otherwise occur.

Legal certainty is also the basis of application of measures such as the Acquired Rights Directive 77/187. This in itself is demonstration of the fact that there is respect for acquired rights that cannot later be withdrawn. This in effect feeds into another aspect of the principle, that there should be protection of legitimate expectations. In simple terms it means that ‘assurances relied on in good faith should be honoured’.

However, it must also be remembered that the institutions are still bound to act in furtherance of the objectives of the Treaties and the principle cannot be employed to frustrate that end.

Natural justice

This is another concept that will be familiar to students of English constitutional law. In fact, the ECJ has on occasions referred to it simply as ‘fairness’.

Within English law there are two distinct strands to the principle:

  • the right to a fair and unbiased hearing; and
  • the right to be heard before the making of a potentially adverse decision is made.

A third aspect that is explicit in many areas of EU law is:

  • the right to a reasoned decision.

The ECJ first addressed the right to a hearing at quite an early stage:

EU law in any case includes many express provisions that guarantee the principle. If Member States choose to claim the derogations under Directive 64/221 (now in Directive 2004/38) as applied to Articles 45, 49 or 56, then they must provide both a proper hearing and a right to appeal. Indeed, this is also the case when Member States make decisions on recognition of qualifications for establishment under Article 43 or provision of services under Article 49: Directives 89/48, 92/51 and the overarching ‘Slim Directive’ 2001/19 (now incorporated in the Qualifications Directive 2005/36) (see Chapter 13). The ECJ has also enforced the principle in the case law on those areas.

Those Directives also guarantee the right to a reasoned decision. The ECJ has also upheld this right in the case law.

The protection of fundamental human rights

There was no mention of human rights in the original EC Treaty. This is actually not all that surprising since the major concern of the initial Treaties was the creation of the Common Market, so it was essentially economic in its direction.

The first statement of the ECJ on the matter came in Stauder v City of Ulm (Case 29/69) [1969] ECR 419. Here, the Court did little more than to confirm that there was nothing in the provision that was being challenged that was ‘capable of prejudicing the fundamental human rights enshrined in the general principles of Community law and protected by the court’.

In Internationale Handelsgesellschaft (1970) the Court was more explicit and more expansive in its statement:

It became apparent that there was a lacuna in the Union legal order in regards to the protection of human rights, as the impact of Union legislation and actions on individual lives grew in importance. If we consider the federalist ambitions of the founding fathers, and the constitutionalisation process of the Union, in order to appear more like a democratic and legitimate political entity, the need to protect human rights, i.e. the rights of both its citizens and the people residing and working across the Union became imperative. Therefore, the absence of a system of checks and balances enabling acts of the Union to be reviewed for breach of individual human rights became a real issue. Consequently, answering the need for a Union approach to human rights, since 1974, the ECJ started affirming and proclaiming in its case law that Union institutions are bound by fundamental principles of basic human rights, despite a lack of specific written provision in the founding Treaties.

The meeting of the European Council at Cologne in 1999 decided that a Charter of Fundamental Rights for the EU should be drawn up in order to provide a more visible means of protection of the citizens of the Union. A Draft Charter has subsequently been produced which was signed and solemnly declared by all of the then 15 Member States at Nice in 2000. As seen in Chapter 1 the CFR has since been given legal binding force since 2009 by Article 6 TEU as amended by the Treaty of Lisbon.

Subsidiarity

Subsidiarity is also not a new concept. In fact, there were references to the principle in the founding Treaties which identified that decisions should be taken as closely as possible to the citizens that are affected by them. In other words, subsidiarity underpins the principle of conferral which is enshrined in Article 5 TEU. This provision essentially requires that, in areas of non-exclusive EU competence, the EU should leave Member States to act in the furtherance of the EU’s objectives wherever appropriate.

Article 5 provided:

In simple terms, then, the principle is that the institutions of the Union should only act to introduce measures where it is more appropriate than for the Member States to act individually. The result is in effect a two-part test:

  • it must be determined that the measure is one which is within the competence of EU law to deal with; and
  • introduction of EU measures can only then be justified if this serves an end which:
    • cannot be achieved satisfactorily at national level; and
    • can be achieved in a more satisfactory way by the EU.

3.3.3 Other tertiary sources

There are certain other tertiary sources of less certain legality. These are of three main types:

  • Acts adopted by representatives of Member State governments meeting in Council (The Council is part of the legislative process under Article 288. However, meetings of representatives of the Member States in the Council are also used to decide on various joint action. This is a quick and easy method of making decisions that fall outside the competence of the EU. While resolutions coming out of such meetings do not have the full force of law, the ECJ will consider them. An example is Commission v Council (Re ERTA). (Case 22/70) [1971] ECR 263 on the legality of an agreement to co-ordinate approaches in negotiation towards a European Road Transport Agreement.)
  • National law of Member States (National law is not formally recognised as being part of EU law. There are, however, two instances when it will be taken into account: first, where Community law actually makes reference to national law, as, for example, in determining the legal status of individuals, i.e. capacity; second, where national law has developed EU law and the ECJ looks to that law for guidance when there is a gap in the law.)
  • International Treaties negotiated by the EU (This refers to multinational Treaties to which the EU is a party. An example would be the General Agreement on Tariffs and Trade (GATT). In International Fruit Co NV v Produktschap voor Groenten en Fruit (Cases 21 and 22/72) [1972] ECR 1219 the ECJ held that GATT could be referred to when determining what practices breach Community (now EU) commercial policy.)

SUMMARY

  • The sources of EU law are of three types: primary, secondary and tertiary.
  • Primary sources are the Treaties themselves and, as well as outlining the objectives and indicating the roles of the institutions and processes, these also contain much substantive law.
  • Secondary sources are the legislation from Article 288 TFEU – the main three types are regulations, directives and decisions.
  • The main tertiary sources are the case law of the Court of Justice and the general principles of law – many of the major principles of EU law, e.g. supremacy – have developed from the case law.

Further reading

Articles

  • Meyring, B, ‘Intergovernmentalism and Supranationality: Two Stereotypes for a Complex Reality’ (1997) 22 ELR 221.
  • Wouters, J, ‘Institutional and Constitutional Challenges for the European Union: Some Reflections in the Light of the Treaty of Nice’ (2001) 26 ELR 342.

Books

  • Craig, P and de Burca, G, EU Law: Text, Cases and Materials (OUP, 2020).
  • Schütze, R, European Union Law (OUP, 2021).
  • Ward, I, A Critical Introduction to European Law (3rd edn, CUP, 2011), chap. 1.

4 The legislative process

DOI: 10.4324/9781003218562-4

4.1 The role of the institutions

The process of legislating within EU law appears to be quite complex. One of the reasons for this is that all rules and procedures for legislating are laid down in the Treaties at different points. Every EU law is based on a specific Treaty Article, in this case referred to as the ‘legal basis’ of the legislation. The second reason is that there are different processes of legislating and the appropriate process depends on the particular area of the Treaty objectives that requires legislating.

The process of legislating has been modified very significantly since the original Treaties, as the result of the different subsequent Treaties. The introduction of new legislative procedures arose from the criticism of the early legislative process that there was a so-called democratic deficit, in other words that those institutions that existed on the basis of appointment rather than election controlled the processes. This democratic deficit was considered to be particularly true of the role of Parliament which originally had little effect on the legislative process other than to suggest amendments. The Lisbon Treaty has modified and simplified the legislative process and now most legislation is created under the co-decision procedure.

The three institutions mainly involved in the legislative process are:

  • The Commission (the body mainly responsible for ‘proposing’ legislation and producing draft legislation; in short it is for the Commission to initiate legislation except where the Treaties provide otherwise).
  • Parliament (is one of the two key bodies responsible for passing legislation in the EU. It can also have either a consultative role or a right of veto under SLP).
  • Council (the second half of the legislative duo with the EP).

Other institutions also have a role in terms of receiving draft proposals for legislation and providing consultation, particularly the Economic and Social Committee and the Committee of the Regions. COREPER, of course, has a role to play in supporting Council through all legislation.

Prior to the Treaty of Lisbon there were basically four types of legislative procedure that were possible within the EU legal order. However, it is also true that there are limited circumstances where the Commission is authorised to legislate on its own and there are other instances where the Council and Commission can act without consulting Parliament, although in practice they still do.

The four main processes were:

  • the proposal (or consultation) procedure;
  • the co-operation procedure (this has now been discontinued);
  • the co-decision procedure (this is now the main legislative process now known as the ordinary legislative procedure under Article 294 TFEU);
  • the process of assent.

The proposal procedure was in fact the original legislative procedure which was used prior to the SEA.

Because of the criticism that there was insufficient accountability, the so-called democratic deficit, a new procedure was introduced in the SEA to give Parliament a greater and more meaningful role. This was the ‘co-operation procedure’. It was introduced to provide a relatively straightforward means of involving the European Parliament, which would have two readings of the draft proposals. It was also based entirely on qualified majority voting by Council. Its main context was for Internal Market measures. Although it was introduced to reduce the democratic deficit and did indeed succeed in giving Parliament a greater level of involvement, the procedure was still criticised:

The TEU then introduced the predominant process nowadays: the ‘co-decision’ procedure, now the ordinary legislative procedure under Article 294 TFEU. A complex process originally, it has subsequently been somewhat modified by the Treaty of Amsterdam for the sake of simplification. Use of the procedure was expanded by the Treaty of Nice and has now become the main legislative process for adopting EU law since the Treaty of Lisbon. The ordinary legislative procedure has put the EP on an equal footing as the Council in the adoption of EU law and has thus contributed to the democratic credential of the EU.

The European Parliament may also participate in the law-making process in special legislative procedure, such as the consent and consultation procedures.

Following the Treaty of Lisbon coming into force, the ordinary legislative procedure (formerly the co-decision procedure) is the major process for legislating. It involves an active role by all three institutions, including Parliament.

As a result of the Treaty of Lisbon, Article 5(1) TEU requires for a process of conferral to limit the competences of the EU to legislate in particular areas. Where the EU institutions are not granted competence in the Treaties then there may be shared competence with the Member States, or the EU may act to support, co-ordinate or supplement the actions of Member States.

Areas where the EU institutions have exclusive competence are: the Customs Union, competition rules, monetary policy for Member States in the euro zone, marine biology and fisheries policy, and international agreements.

4.2 The ordinary legislative procedure (formerly the co-decision procedure)

The TEU introduced the process known as the ‘co-decision procedure’, whereby the European Parliament and Council must both agree on a legislative proposal before it can become law – in some way, it is like a bicameral structure to the EU legislative process. The aim was to give far greater power to Parliament by allowing it not only to suggest significant amendments but also ultimately to have some right of veto on draft legislation. In this respect it can be said that Council and Parliament share legislative power. The process was amended and simplified by the Treaty of Amsterdam and to a lesser extent by the Treaty of Nice. Following the Treaty of Lisbon this is now the ordinary legislative procedure used for most legislation. The ordinary legislative procedure is enshrined in Article 289 TFEU which states it ‘shall consist in the joint adoption by the European Parliament and the Council of a regulation, directive or decision on a proposal from the Commission’. The procedure for adopting legislation under the ordinary legislative procedure is laid down in Article 294 TFEU.

Parliament gains more power because after it proposes amendments Council must then consider them before it acts, and if Parliament rejects the proposals then the act is not adopted. Both Council and Parliament have two readings of the proposal, and if they cannot agree then the measure is put before a ‘conciliation committee’ which is made up of members from both institutions in equal numbers. Both bodies then hold a third reading so that they may finally adopt the proposal as law.

After the first consultation of Parliament Council may adopt the proposal by Qualified Majority Vote if either:

  • it agrees with all of Parliament’s amendments; or
  • Parliament has not made any amendments.

If this is not the case then Council adopts a ‘common position’ by Qualified Majority Vote which is then communicated to Parliament. Within three months if Parliament either approves the common position or does nothing then Council can adopt the common position.

Parliament can, however, within this three-month period:

  • reject the common position by absolute majority of MEPs – in which case this acts like a veto and the measure cannot be adopted; or
  • propose new amendments by an absolute majority.

If new amendments are introduced these are forwarded to the Commission which itself has three choices:

  • it can accept them all; or
  • it can reject them all; or
  • it can selectively accept some and reject others.

On receiving Parliament’s amendments, Council also has choices:

  • it can approve them all and adopt the amended measure by Qualified Majority Vote; or
  • it can fail to adopt the proposal and convene a ‘Conciliation Committee’.

Within six weeks the committee then has two choices:

  • it can approve a new joint proposal decided upon by the committee which can then be adopted by Parliament by absolute majority and Council by qualified majority; or
  • it can fail to find any possible compromise, in which case the measure is not adopted.

The procedure can be explained in diagram form as in Figure 4.2.

A flow diagram depicts the different stages in the ordinary legislative procedure, with three institutions involved in this process: the Commission, Parliament, and Council.

Long Description for Figure 4.1

The Commission produces a draft proposal and sends it to Parliament and the Council. The Parliament delivers its opinion after first and second readings to the Council and approves or rejects the common position based on the Council’s amendments or no amendments. The Commission gives a positive or negative opinion of the Parliament’s amendments, and then in the Council, the measure is adopted or not adopted based on it. If the measure is not adopted, then it is sent to the conciliation committee. It may be approved or fail to be approved and sent to the Council and Parliament.

Figure 4.1 The different stages in the ordinary legislative procedure (formerly the co-decision procedure)

A flow diagram depicts the different stages in the proposal, with three institutions involved in this process: the Commission, Parliament, and Council.

Long Description for Figure 4.2

The Council passes laws after consultation with Parliament and the Economic and Social Committee. Both send the opinion to the Commission and Committee of Permanent Representatives. The Commission sends the original and amended proposals to the Committee of Permanent Representatives.

Figure 4.2 The different stages in the proposal (consultation) procedure

The development of the procedure through both the Treaty of Amsterdam and the Treaty of Nice means that the procedure is now used for most areas of legislation. These now include:

  • non-discrimination on the basis of nationality;
  • the right to move and to reside;
  • the free movement of workers and social security for migrant workers;
  • the right of establishment;
  • transport;
  • employment;
  • Customs co-operation;
  • the fight against social exclusion;
  • equal opportunities and equal treatment;
  • implementing decisions involving the European Social Fund or the Regional
  • Development Fund;
  • education and vocational training;
  • culture;
  • health;
  • consumer protection;
  • trans-European networks;
  • research;
  • the environment;
  • transparency;
  • the prevention of fraud;
  • statistics;
  • the setting up of a data protection advisory body.

4.3 Special legislative procedure

Article 289(2) TFEU states that ‘in the specific cases provided for by the Treaties, the adoption of a regulation, directive or decision by the European Parliament with the participation of the Council, or by the latter with the participation of the European Parliament, shall constitute a special legislative procedure’. The adoption of the special legislative procedure is stipulated by specific Treaty provisions and is therefore used on an ad hoc basis.

One form of special legislative procedure is the consent procedure, formerly known as the ‘assent’ procedure, which gives Parliament the power to veto any legislative proposal. Under this procedure the EP can vote to either accept or reject a proposal, but cannot amend it. Failure to secure the EP’s consent means that a given proposal cannot be adopted into law.

By contrast, under the other special legislative procedure known as the consultation procedure, the EP participates in the law-making process in a purely consultative capacity. Whilst the Council is bound to seek the opinion of the EP, it is however not bound to follow it.

This was the original process used for all legislation. Following the Treaty of Nice, it was still the basis for the adoption of certain general EC (now EU) instruments or policy areas. Examples are in Article 64(3) TFEU and Article 86(1) TFEU.

This procedure most accurately represents the notion that it is the Commission that ‘proposes’ and the Council that ‘disposes’.

Nevertheless, the process still depends on some consultation with Parliament as well as with the Economic and Social Committee or the Committee of the Regions, and discussions in COREPER. Failure genuinely to consult may amount to a breach of an essential procedural requirement. On this basis it may result in the measure being declared void.

The importance of consultation was recognised by the ECJ in Roquette Frères SA v Council (Case 138/79) [1980] ECR 3333.

The process involves the Commission sending its proposal to both Council and Parliament, as well as one of the committees if appropriate. The measure cannot then become law until Parliament has delivered its opinion. Even so, Council has the final say.

Whilst the above was one more justification that there was a democratic deficit in the administration and law-making of the EC, this assertion may lack some of its sting since the institutional reforms brought on by the Treaty of Lisbon, especially with regards to the increased role of Parliament.

The most significant difference between the various legislative procedures is the involvement and therefore democratic influence of Parliament in the law-making process. Under the consultation procedure Parliament only gives its opinion on draft legislation, so under this procedure the Council is the main law-maker. By contrast, under the consent procedure, the power of veto of the EP makes it the ultimate law-maker under this special legislative procedure. However, the balance of power is equalised under the ordinary legislative procedure where both the EP and the Council are co-legislators and share the power to make law in the EU.

SUMMARY

  • Originally most legislation was introduced through the proposal procedure which had a very limited role for Parliament.
  • Complaints of a ‘democratic deficit’ led to the introduction of the co-operation procedure, which involved greater consultation of Parliament but still little power to influence legislation.
  • Following the ToA, and with modification by ToN, the co-decision procedure was introduced – this is now known as the Ordinary Legislative Procedure since the Treaty of Lisbon, it is used for most legislation and gives Parliament greater powers to make amendments and influence their outcome.

Further reading

Article

  • Dashwood, A, ‘The Constitution of the European Union after Nice: Law-making Procedures’ (2001) 26 ELR 215.

Books

  • Craig, P and de Burca, G, EU Law: Text, Cases and Materials (OUP, 2020).
  • Douglas-Scott, S, Constitutional Law of the European Union (Longman, 2002), chap. 3.
  • Schütze, R, European Union Law (OUP, 2021).

5 Enforcement of EU law (through ‘direct’ and ‘indirect’ actions)

DOI: 10.4324/9781003218562-5

5.1 The nature and purpose of enforcement

‘Enforcement’ in simple terms refers to the actions created in the EC Treaty (now TFEU) for the purpose of ensuring that both the Member States and the institutions of the EU comply with their relative obligations within the Treaties.

The Treaties and their secondary legislation clearly create many and various substantive rights and obligations by which all parties to the Treaties are bound. These substantive rights and obligations granted under Treaties would, nevertheless, be left completely ineffective if they were left merely to the co-operation of Member States without the means of enforcing them.

Similarly, it would also be possible that the individual rights might be abused by the institutions of the EU themselves.

For these reasons the framers of the EC Treaty were wise enough to include a variety of enforcement proceedings, and the methods for reviewing the actions of both the institutions and the Member States. These were then placed under the scrutiny of the ECJ, with individuals able to gain remedies following actions in their favour.

These procedures are generally referred to as the ‘direct actions’. They supplement the ‘indirect actions’ of the Article 267 reference procedure.

The measures are broadly based in that they allow a wide range of applicants to take the initiative in setting an action in motion and bringing proceedings. So this might include, for example, other institutions of the EU, as well as private citizens in certain circumstances.

There are essentially four types of action to be considered here:

  • Actions under Article 258 TFEU (formerly Article 226 EC) are taken by the Commission against Member States for a failure to honour their Treaty obligations (and generally known as ‘infringement proceedings’).
  • Actions under Article 263 TFEU (formerly Article 230 EC) can be brought against any institutions of the EU for acting in excess of their actual powers – since the powers of the various institutions are defined in the Treaties, and since, in legislating, the institutions are only capable of acting for furtherance of the actual objectives of the Treaties. This is more commonly known as the judicial review procedure.
  • Actions under Article 265 TFEU (formerly Article 232 EC) are taken against the institutions of the EU where they have failed to act when they are required to act (hence it is known as an action for failure to act).
  • Actions under Article 340 TFEU (formerly Article 288 EC) is an action for damages brought against the institutions to compensate a citizen for any loss caused by one of the institutions, so it is inevitably linked with grievances pursued under the previous two.

5.2 Indirect actions – Article 258 infringement proceedings against Member States

5.2.1 Actions against Member States under Articles 258 and 259

It is in the very nature of EU law that it depends on a partnership between the Member States and the institutions. This can be seen, for instance, in the implementation of Directives where the legislation is created under the various processes but in the form of a written obligation containing various objectives that the Member States must incorporate into their national law but in a manner of their choosing and within an implementation period set by the Council.

While this partnership exists, it is not uncommon for Member States to show carelessness in implementing them (e.g. this was the case in the deficiencies in UK sex discrimination law in relation to the issue of different retirement ages for men and women based on state pension ages highlighted in the case of Marshall v Southampton and South West AHA (Case 152/84) [1986] QB 401). It is even possible that Member States may show real reluctance in fulfilling their obligations. (An obvious example of this is the arguments presented by the Tory government before 1997 in refusing to implement the Working Time Directive, despite its argument of an opt-out applying to the provision having been rejected out of hand by the ECJ.)

It is because of these possibilities that the Treaty sensibly provided the means of calling Member States to account for their failures to honour Treaty and legislative obligations. The process can be initiated in one of two ways:

  • in most cases the process would normally be invoked by the Commission under Article 258;
  • however, it is also possible for proceedings to be initiated by other Member States under Article 259.

5.2.2 Actions by the Commission under Article 258

The Commission has always been described as the ‘watchdog of the Treaties’ and so it was empowered by Article 258 to act in such manner monitoring the behaviour of Member States and, if necessary, enforcing compliance with Treaty and other obligations. It therefore has locus standi in Article 258 actions.

Article 258 itself provides the following:

As a result, it is possible to identify three clear purposes of the Article 258 action:

  • to ensure that Member States comply with their Treaty obligations;
  • to provide a procedure for resolution of disputes between the Commission and Member States (it must be noted in this respect that four-fifths of disputes are actually settled within the preliminary stages);
  • where proceedings do nevertheless end up in the CJEU, the action also provides a means of clarifying the law for all Member States to follow in the future.

There are in fact three formal stages in the procedure. However, these are usually preceded by an informal stage.

The informal stage

Mediation

Once the possibility of non-compliance has been notified, the Commission at first engages in informal discussions with the Member State. The Commission identifies the nature of the breach by the Member State and will prescribe a time limit within which it expects the Member State to comply. Usually, the Member State is prepared to remedy its mistake at this point, as a result of which the action is generally then suspended.

Formal stages

Formal notice of default

It may be that the Commission is dissatisfied with the response of the Member State. If this is the case then the Commission issues a notice inviting the Member State to submit its own observations on the alleged non-compliance.

It is this stage that in effect defines exactly what the failure by the Member State is and therefore also the terms of reference of the action. These are then fixed and the Commission cannot afterwards extend the scope of the action.

Reasoned opinion

If the notice of default fails to force the Member State into remedial action and it has still not complied with its obligations then the Commission issues a reasoned opinion.

This formal document sets out all the reasons why the Commission considers that the Member State is in default. The reasoned opinion also sets a time limit within which the Commission expects the Member State to act.

Nevertheless, the reasoned opinion is not a binding act in its own right, as a result of which action in the Court of Justice may still be necessary.

Court proceedings in the CJEU

Court proceedings depend on the action taken by the Member State following the earlier proceedings. If the Member State still fails to comply even after the other stages then the Commission will bring an action in the CJEU.

Nevertheless, it is still possible for the action to be settled even without a decision of the Court. Interim relief under Article 243 is an example of this. In fact, something in the region of 44% of cases are settled at this point without further court action.

There are, of course, many defences that Member States have attempted to use, but most have failed:

  • Member States have tried to claim that internal difficulties have genuinely prevented them from meeting their obligations:
  • Reciprocity has also been argued – that national compliance is dependent on compliance by the other Member States:
  • Member States have also argued the application of force majeure – that they are excused from acting when the circumstances are beyond the control of the national authorities:
  • Failing to act because of internal political difficulties such as objections by trade unions has also been rejected as a defence:
  • Member States have also tried to justify breaches of EU law on ethical grounds:
Enforcement

Before the signing of the Maastricht Treaty (TEU), decisions made by the CJEU lacked the possibility of actual enforcement. Because of this, repeated failure by a Member State to comply would simply lead to further Article 258 proceedings.

Following the TEU it was possible under Article 258 for a financial penalty to be imposed on the Member State. An example of this is in Commission v Greece (Case C–387/97) [2000] ECR I–5047 which concerned the dumping of toxic waste. The Court of Justice applied a penalty, taking into account guidelines issued by the Commission. The penalty is based on a basic sum for every day of the breach, multiplied by other significant factors such as the seriousness and length of the breach, and also takes into account the ability of the Member State to pay. The Lisbon Treaty has made procedural improvements to the enforcement procedure through Article 260 TFEU, although in merely stipulating that Member States ‘are required to take the necessary measure to comply with a judgment of the Court’ under Article 258, some have argued that these rulings mostly have declaratory force. Nevertheless, the imposition of pecuniary penalties for a Member State’s non-compliance with obligations under EU law is still a persuasive enforcement mechanism.

Now, where a Directive has not been implemented, the CJEU can apply a penalty against a Member State at the same time that it gives its ruling. Also, in the case of other breaches, it does not have to provide a reasoned opinion.

It is possible for a Member State to present a defence against an action under Article 258 TFEU, whether on procedural grounds (i.e. for failure to comply with the procedural requirements laid down in Article 258) or on substantive grounds (e.g. unlawful obligation, force majeure, uncertain meaning of the obligation).

5.2.3 Actions by other Member States under Article 259

It was always intended in the Treaty that the Commission should be the prime body in enforcing EU law against Member States that were failing to comply with their obligations. For this reason also, while it is possible for other Member States to bring action, it was always intended to be an exceptional procedure rather than the norm. This is reflected in the wording of the Article:

However, the ability of other Member States to use the procedure is nevertheless a very useful safeguard against possible errors of judgement by the Commission.

It involves similar processes to those that we have looked at above. However, the Article clearly demands that the Member State must inevitably work closely with the Commission in the preliminary stages.

To date, whilst there have been four cases heard before the ECJ, only one has actually been successful:

Nevertheless, the possibility of such action can be useful in focusing the attention of the Commission on the issue. This was the case when the EU lifted its ban on British beef but both the French and German governments declined to do so.

5.3 Direct actions against EU institutions

5.3.1 Article 263 TFEU actions against EU institutions for exceeding powers

The action for annulment of an EU instrument under Article 263

Article 263 provides one of the few circumstances in which ordinary individuals are able to bring an action in the CJEU. While such an action is possible, it is nevertheless true that the ability of citizens to do so is much more restricted than it is for the institutions.

The procedure is a very specific one and it has two major functions:

  • it provides a means of questioning and indeed controlling the legality of binding acts of EU institutions;
  • it offers a form of legal protection to those who are subject to the instruments of the EU and who are adversely affected by instruments that are in fact illegal.

There are three key aspects to Article 263 actions that must be considered:

  • the identity of those who may bring an action, in other words the locus standi of individuals;
  • the type of actions by the institutions that are capable of being reviewed under the procedure and the grounds on which an action may be brought;
  • the actual procedure itself.

Locus standi (the right to sue)

Article 263 is quite explicit on those who have locus standi. There are three significant groups enjoying slightly different locus standi:

  • The Member States, the Commission and the Council are all named as possible parties to an action. In this respect they are all classed as ‘privileged claimants’ and possess virtually unlimited rights of challenge against any act of any of the institutions. The exception to this is recommendations and opinions. The necessary requirement for being a privileged claimant is that the body is bound by the measure in question. This will be determined by the CJEU on studying the context and legal effect.
  • Parliament and the European Central Bank are also privileged claimants. However, they have more limited powers of challenge. Traditionally, it was held that they could only use the procedure if it was ‘for the purpose of protecting their prerogatives’. An example of this is Parliament v Council (Chernobyl) (Case 70/88) [1991] ECR I–4529. The Treaty of Nice, however, has identified Parliament as having full status as a privileged claimant.
  • Natural and legal persons are also identified in Article 263 as having locus standi. Their rights of challenge before the Treaty of Lisbon were initially limited to ‘a Decision addressed to that person, or a decision which, although in the form of a Regulation or a Decision addressed to another person, is of direct and individual concern to the individual’. Since the Treaty of Lisbon reforms, Article 263(4) TFEU extends the challenges available to individuals to include:
    • ‘an act addressed to that person’ (e.g. a decision)
    • ‘or which is of direct and individual concern to them’ (i.e. a legislative act)
    • ‘and against a regulatory act which is of direct concern to them and does not entail implementing measures’

The distinction between a legislative act and a regulatory act is essential in that it will determine the judicial test which the ECJ will apply to establish whether a natural/legal person has locus standi under Article 263 TFEU.

Whilst the threshold for challenging a legislative act is cumbersome (the non-privilege applicant (NPA) must prove both individual and direct concern), the threshold for challenging a regulatory is less taxing as the NPA only has to fulfil the criteria for direct concern.

The distinction is made with regards to the procedure through which a legal act of the EU is adopted. If, according to Article 289(3) TFEU, a legal act is adopted by a legislative procedure, whether the ordinary legislative procedure or special legislative procedure, then the measure is a legislative act. This was confirmed in Inuit Tapiriit Kanatami (ITK) (Case C–583/11 P) [2014] QB 648 in which the ECJ stated that acts of the EU institutions that are legally binding and which cover ‘acts of general application, legislative or otherwise, and individual acts’ can only be challenged by NPA if they show individual and direct concern. It follows then that regulatory acts ‘must be understood as covering all acts of general application apart from legislative acts’; in other words, acts which are not adopted through a legislative procedure, but are instead adopted by the European Commission through delegated powers (e.g. Article 290 TFEU) or for the exercise of implementing powers (Article 291(2) TFEU). This was confirmed in Microban (Case T–262/10) [2011] ECR II–7697.

All challenges by natural and legal persons are now brought in the General Court. There is the possibility of an appeal to the CJEU but on a point of law only.

Looking at the requirements above, it is clear that, apart from in the case of decisions addressed to an individual, three key issues need to be considered in establishing whether or not there is locus standi:

  • what amounts to ‘individual concern’;
  • what amounts to ‘direct concern’;
  • the circumstances in which a Regulation may be of individual or direct concern.
Individual concern

In order for a private applicant to make a claim, ‘individual concern’ must mean that the Decision or Regulation must affect the applicant. How an applicant could claim to be affected by the measure was explained in Plaumann v Commission (Case 25/62) [1963] ECR 95.

However, this basic test has subsequently been modified so that now it must be possible to determine the number and the identity of those persons affected at the time that the measure complained about was adopted.

This has also been modified further by the ECJ which, in International Fruit Co v Commission (Cases 41 to 44/70) [1971] ECR 411, has stated that there will be individual concern and therefore locus standi is possible if there is a ‘closed group’ of people affected by the decision. In the case there was such a ‘closed group’ because the decision only applied to a limited number of importers who had been granted licences before a specific date.

However, the Court of Justice has not always shown consistent application of the criteria that it has set:

Nevertheless, the definition of individual concern in Plaumann (1963) is restrictive and makes it difficult for individuals to protect themselves against breaches of their rights resulting from EU legislation. The test has been consistently criticised. The definition has recently been reviewed and a more liberal approach suggested. However, the Court of Justice has ultimately confirmed Plaumann as the appropriate test.

In Union de Pequenos Agricultores (UPA) v Council (Case C–50/00) [2003] QB 893 UPA, a trade association, had unsuccessfully challenged a Regulation in the CFI (now the General Court), being unable to show individual concern. The CFI (now the General Court) had also pointed out that UPA could instead have brought an action in the national courts and then asked for an Article 267 reference to be made. When the case came before the ECJ the Advocate-General identified that a challenge under Article 263 was a more appropriate procedure and recognised that there were inherent difficulties in trying to take the course of action suggested by the CFI (now the General Court). First, a national court would not have the power to annul the measure and so could only consider whether there was sufficient doubt as to its legality to justify a reference being made. Second, certain measures could not give rise to an action in a national court and so would be beyond any challenge by the individual. He also felt that the definition of ‘individual concern’ was too restrictive and that there was no reason why an individual should have to show a difference from other individuals affected by the measure. He preferred a test based on an individual having suffered a substantial adverse effect because of his particular circumstances.

Between the Advocate-General’s opinion and the ruling in the ECJ, the CFI in another case, Jego-Quere et Cie v Commission (Case T–177/01) [2003] QB 854, suggested a different test for individual concern based on the Advocate-General’s opinion in UPA. Individual concern would be shown if the measure ‘affects his legal position in a manner which is both definite and immediate, by restricting his rights or imposing obligations on him’.

The ECJ in UPA, however, confirmed the Plaumann (1963) test on individual concern so that there is unlikely to be change without amendment to the Treaty.

Direct concern

‘Direct concern’ has a somewhat different meaning, and once again has been subject to some inconsistent interpretation by the Court of Justice.

It does not only refer to the causal connection between the decision and any loss suffered but has also been said to refer to the ‘immediate, automatic and inevitable disadvantageous legal effects’ without need for further intervention.

However, this somewhat strict approach was later relaxed in Bock v Commission (Case 62/70) (The Chinese Mushrooms Case) [1971] ECR 897, where the applicant who had applied for a licence to import Chinese mushrooms was granted locus standi because only he had been affected.

When a Regulation is of individual or direct concern

Because of their nature genuine Regulations can never be capable of challenge by an individual applicant. This was stated clearly in Calpak SpA v Commission (Case 789/79) [1980] ECR 1949.

Because of this it is vital to determine whether or not a particular Regulation conforms to the standard definition, otherwise there can be no locus standi in any application challenging it. In Confederation Nationale des Producteurs de Fruits et Legumes v Council (Cases 16 and 17/62) [1962] ECR 901 Advocate-General Lagrange identified:

In this way a Regulation can only be challenged when it is not a provision having general application within the meaning given in Article 288 but is rather ‘a bundle of individual Decisions taken by the Commission, each of which, although taken in the form of a Regulation, affected the legal position of the applicant’ (International Fruit Co v Commission (1971)).

The substantive grounds for review

Once admissibility of an application has been established then it is for the applicant to show that the challenge to the decision concerns one of four specific grounds identified in the Article:

  • lack of competence;
  • infringement of an essential procedural requirement;
  • infringement of the Treaties or of any rule that relates to the application of the provision of the Treaties;
  • misuse of power.
Lack of competence

This ground for complaint has no real comparison in English administrative law. However, Lasok and Bridges have suggested that it is ‘broadly comparable’ with the ultra vires doctrine with which we are all familiar.

In simple terms it is possible to identify lack of competence when an EU institution appears to exercise a power that is not in fact conferred upon it by EU law; or where it appears to exercise a non-existent power; or where it in fact encroaches on the power given to another institution.

The ECJ has defined the ground in the case law but it will rarely accept a challenge by one institution against another because, first, the powers of the different institutions are clearly laid out in the Treaty and, second, because it will in any case usually interpret these powers broadly.

In this way it is more likely for the ground to be used in respect of powers that are not possessed at all by the institution challenged in the application.

Most commonly, the ground will be accepted in the case of an improper delegation of power. This was the case in Meroni v High Authority (Case 9/56) [1956–58] ECR 133 where the Commission (High Authority) had delegated powers to make a decision to another body that did not in fact have any authority to make decisions. The action of the Commission in this instance was invalid.

Infringement of an essential procedural requirement

EU law puts in place a number of procedural mechanisms in order to act as safeguards in protection of natural justice. Such essential procedural requirements fall into distinct categories:

  • First, there are procedural requirements in relation to the preparation of the measure, for instance the requirement of prior consultation. For example, in Roquette Frères v Council (Case 138/79) [1980] ECR 3333 the Council had failed to consult Parliament, as required, on agricultural budgeting measures and so the measure was invalid.
  • Also, there are requirements in respect of the form by which the measure is created. An example of this would be the requirement to give reasons so that any party affected by the measure can understand how the institution has applied the law. Such an infringement occurred in Germany v Commission (Case 24/62) [1963] ECR 63 where a decision addressed to Germany concerning wine imports provided no reasons and so was invalid.
Infringement of the Treaties or of any rule relating to their application

This ground is easily explained. It clearly allows the CJEU to review how the acts of the institution in question conform with EU law. The law in this instance includes the general principles of law so that any kind of violation of EU law of whatever type may be declared invalid under this ground.

Misuse of powers

This ground quite simply refers to the situation where an institution is using a power that it does in fact possess but for an objective that is contrary to those for which the power was given. It might therefore include any illegitimate use of a power.

The time limit of the procedure

The most important procedural requirement in making an application is the existence of a strict time limit for bringing an action. According to Article 263(6), the time limit is within two months of the date on which the measure was published, or from the date on which the applicant was notified of the decision, or on which it came to the applicant’s attention.

However, Article 50 of the Rules of Procedures of the Court of Justice (RPCJ) (L 265/1) further provides that the actual time limit for proceedings against a measure adopted by an institution will be calculated from the end of the 14th day after publication. Therefore, the countdown would technically only start 15 days after the Regulation was published in the OJ. Taking into account distances which may hinder or slow down the process of publication or notification, NPAs may also benefit from an additional 10 days (Article 51 RPCJ). Finally, the Court of Justice may also decide to extend the time limit allocated to the applicant (Article 52 RPCJ).

Nevertheless, the time limit criterion is generally strictly applied. Derogations could only apply where unforeseeable events, force majeure or excusable errors have delayed the submission of the claim for annulment (Transportes Evaristo Molina (Case C–36/09 P) [2010] ECR I–145); or where an EU measure is deemed defective if, for instance, its legal basis or the rights/obligation it creates are unascertainable. The ECJ held that ‘such a measure produces no legal effects and may be challenged outside the limitation periods’ (BASFAG & Others v Commission of the European Communities (Joined Cases T–79, 84–86, 89, 91, 92, 94, 96, 98, 102 and 104/89) [1992] ECR II–315).

The effects of a successful application

Article 263 is essentially an action for the annulment of an EU measure which infringes a substantive or procedural rule of law. The procedure under Article 263 is thus essentially concerned with first establishing the admissibility of a claim from the applicant (especially NPA), and second to determine whether the disputed EU measure is in fact illegal.

If the ECJ finds that the measure is indeed legally defective, the claim is successful and the outcome is stipulated by Article 264 TFEU which provides that such a measure shall be declared void. The effect of this is that the measure is treated as though it never existed in the first place.

A defunct measure becomes void, wholly or partially, upon the delivery of the ECJ ruling (ex tunc). The ruling declaring the annulment applies to everyone who might be concerned or affected by that measure (erga omnes), as confirmed in Commission v AssiDomän (Case C–310/97 P) [1999] ECR I–5363 and may also have retroactive effects (Corus v Commission (Case T–171/99) [2001] ECR II–2967).

The consequence of an application being successful is that the instrument is declared void by the CJEU. The effect of this is that the measure is treated as though it never in fact existed.

A flow diagram depicts the requirements for an action for annulment under Article two hundred and sixty-three with “yes” or “no” conditions.

Long Description for Figure 5.1

The flow begins with the question, “Does the party seeking annulment have locus standi under Article two hundred and sixty-three “. If ‘yes’, then “Is the measure of direct and individual concern to the individuals?”. If ‘yes’, then “Is the ground on which the measure is challenged”. If ‘yes,’ then “A review is possible under Article two hundred and sixty-three “. All the ‘no’ conditions lead to “No review is possible under Article two hundred and sixty-three “.

Figure 5.1 The requirements for an action for annulment under Article 263

5.3.2 Article 265 actions against institutions for a failure to act

Article 265 gives both the Member States and the EU institutions the right to call to account any of the Council, the Commission, Parliament and the European Central Bank for failing to take action when they would be required to. This is an obviously necessary addition to the annulment proceedings in Article 263. Just as there are times when the institutions go beyond the powers given to them by the Treaties, then there is also the possibility that in situations where one of the institutions would be bound to act according to the law, it fails to act. A classic example is where the Commission fails to issue a decision following a breach of competition law under Article 101 or Article 102. In such situations the Treaty has provided a form of redress to the injured party through the Article 265 action. Applicants are required to satisfy a test for admissibility and also to show that there are suitable grounds for the review.

Admissibility

There is no set time limit for a claim under Article 265 but in order for the Court to accept that there is an admissible claim it will first see whether three conditions are met:

  1. First, the applicant must be able to show locus standi. The ‘privileged claimants’ in this case are described in the Article as ‘the member states and other institutions’. This includes Parliament since Parliament v Council (Case 377/87) [1988] ECR 4051. Following the TEU it also includes the European Central Bank. ‘Natural and legal persons’ are also able to bring an action against a Community institution which failed to address to that person any instrument other than a recommendation or an opinion. Inevitably, this means that the instrument must be a binding act. It also obviously means that there would have been a requirement to address the instrument to the applicant had it been enacted.
  2. Second, there must be what is referred to as an ‘Indictable Institution’. The institutions covered by the Article are the Council, the Commission, Parliament and the Central Bank. For the claim to be admissible there must have been a failure to act by that institution in circumstances where it was in fact legally required to act.
  3. Third, there must have been a prior approach to the institution. Before the CJEU will consider an application it must be satisfied that the applicant has already approached the institution seeking redress. This approach must be explicit in its terms and refer to the possibility of a challenge under Article 265 if no reply is received within two months.

Grounds for review

The grounds under which the CJEU will undertake a review are where the applicant is able to show that they were entitled to a decision and none was actually addressed to them. An alternative ground is where an action has not been taken which is of direct and individual concern to them.

In fact, few cases are found to be admissible and so there are few guidelines. Generally, if there is a result to be achieved and an obligation is sufficiently well defined then any attempt to disregard it will fall within the scope of Article 265.

5.3.3 Article 340 TFEU actions against institutions for damages

Article 340 in para 2 states:

In the case of non-contractual liability the EU shall, in accordance with the general principles common to the laws of the member states, make good any damage caused by the institutions or by its servants in the performance of their duties.

It is possible, then, to see the similarity between the Article 340 action and a form of general tort action. However, it should be remembered that five of the original six members of the EC had forms of civil liability based on the French Civil Code. As a result, liability under Article 340 is more accurately seen as based on this form of liability.

The Treaty in any case provides that the CJEU should hear all actions under Article 340. There are two significant issues to consider:

  • the requirements for admissibility of claims; and
  • the conditions for liability.

Admissibility

Locus standi in such claims is necessarily almost unrestricted. As a result, any natural or legal person is capable of bringing an action. The key requirement for admissibility is that the individual making the claim can make out a prima facie case that he has personally suffered damage resulting from an act or omission of an institution or of its servant. In this way the action could not, for instance, be brought by a trade union on behalf of a member or members.

The ECJ ruled in Werhahn Hansamuhle v Council (Cases 63 to 69/72) [1973] ECR 1229 that the claim must be alleged specifically against an institution or its servant. Therefore a claim could not be made against the EU as a whole.

The appropriate time limit on claiming is five years from the date of the event that it is alleged caused the damage and gave rise to the action.

Conditions for liability

There are three elements that must be satisfied for a successful claim to be made: First, the occurrence of damage suffered by the applicant:

  • This might include any physical damage as well as economic loss, including both actual damage and loss of earnings. The general qualification is that the damage is certain, provable and quantifiable.
  • Future loss is also recoverable but only in limited circumstances. In Kampffmeyer et al. v Commission (Cases 5, 7 and 13 to 24/66) [1967] ECR 245 a claim for a future loss was accepted where the cancellation of contracts had already occurred by the time of the creation of the wrongful measure.
  • Even highly speculative and non-material loss has been accepted in some circumstances:

Second, the presence of fault on the part of the institution complained about:

  • It is sufficient in this sense to show that the applicant was owed a duty which was then breached by the institution, as was the case in Adams v Commission (1986).
  • However, the CJEU may be less likely to conclude that there is fault where the institution was involved in making policy decisions and made errors of judgement leading on to the damage suffered.

Third, it must be possible to show a causal connection between the measure complained of and the damage suffered:

  • On this basis the mere existence of damage by itself is insufficient to give rise to an action for damages under Article 340.
  • Proof of damage alone is insufficient for liability without proof also that the act of the institution challenged directly caused the damage.
  • In this way remoteness of damage is clearly an important factor to be taken into account:

SUMMARY

  • Other than references for preliminary rulings there are four types of action in the Court of Justice – infringement proceedings under Article 258, judicial review of acts by EU institutions under Article 263, and for a failure to act under Article 265, and damages claims under Article 340.
  • Infringement proceedings are against Member States for breaches of EU law and penalties are possible.
  • Judicial review of abuse of powers by EU institutions have two major functions: to ensure the legality of binding acts, and to give legal protection to those subject to EU instruments who are adversely affected by illegal ones.
  • Actions against EU institutions for a failure to act allow privileged claimants and natural and legal persons to claim where the institution has failed to act, for instance where a decision should have been issued.
  • Claims for damages against EU institutions allow for making good any damage caused by those institutions.

Further reading

Books

  • Craig, P and de Burca, G, EU Law: Text, Cases and Materials (OUP, 2020).
  • Kaczarowska-Ireland, A, European Union Law (Routledge, 2016), chap. 15.
  • Schütze, R, European Union Law (OUP, 2021).

6 Article 267 TFEU and the preliminary reference procedure

DOI: 10.4324/9781003218562-6

6.1 The relation with Member States

The Article 267 procedure allows any national court or tribunal in any of the Member States to request that the ECJ interpret provisions of EU law. It is crucial to remember that the ECJ simply interprets EU law: the national court then has the task of applying that law, as interpreted. As Lord Denning explained in the Court of Appeal in Bulmer v Bollinger [1974] Ch 401:

Since Brexit, the UK is no longer a Member State and hence judges in UK courts have lost the necessity or even the ability to seek preliminary ruling. However, Lord Denning’s words still apply today; simply substitute (say) ‘French’ or ‘Hungarian’ or ‘Swedish’ instead of ‘English’ in the above quotation.

It is crucial to appreciate that Article 267 is not an appeal procedure. It is triggered by national courts or tribunals during the course of litigation itself. This is why it is known as the ‘preliminary reference’ procedure, and the Court’s judgments are known as ‘preliminary rulings’. The rulings are designed to assist the national court or tribunal to reach a final ruling.

There is, therefore, a shared jurisdiction between the national courts and the ECJ. The national courts decide questions of fact and national law; it is also the national courts who apply national and EU law. The ECJ determines abstract questions of the interpretation of EU legislation only (and deals with issues involving the validity of EU secondary legislation). In the first ever preliminary reference case, De Geus v Robert Bosch (Case 13/61) [1962] ECR 45, Advocate-General Lagrange said that the ‘provisions of [Article 267] must lead to a real and fruitful collaboration between the municipal courts and the Court of Justice with mutual respect for their respective jurisdiction’.

More recently, in Gintec (Case C–374/05) [2007] ECR I–9517, Advocate-General Ruiz-Jarabo Colomer offered the following culinary metaphor as a means of explaining the operation of the preliminary rulings procedure:

6.2 The character of the reference procedure

6.2.1 References seeking interpretation of EU law

If a dispute as to the proper interpretation of a provision of EU legislation arises during a legal dispute before a court or tribunal in one of the Member States, a request may be made for a ruling on the interpretation of the disputed provision. The national court or tribunal suspends the case until the ECJ gives its ruling. When the ECJ has made its decision, the national court or tribunal then continues from where it left off, applying the EU law as interpreted by the ECJ.

This is supposed to achieve uniformity or consistency of interpretation of all EU law, because once the ECJ has made its preliminary ruling, this establishes a precedent for all the courts and tribunals in the Member States to follow in future cases. Having one court to interpret all EU law means that the same meaning is given throughout the Union; if it was left to national courts they might all come up with different interpretations. This would be likely given that the EU presently has 24 official languages. The Treaties are reproduced in all of them. Given that translation is not a precise science, there are bound to be differences between all the different versions; but using the ECJ helps keep the differences to a minimum. If the preliminary reference procedure did not exist, courts and tribunals in the Member States would have to make their own interpretations of EU law. This would create a very real risk of divergent meanings being given to the same provisions of EU law in different Member States and, if that happened, the whole fabric of EU law could begin to unravel.

The Court can give interpretations of provisions in both of the Treaties, and ‘acts of the institutions, bodies, offices or agencies of the Union’ – which basically means all secondary EU legislation (primarily Regulations and Directives). It can also rule on the interpretation of international treaties entered into by those institutions (Hageman (Case 181/73) [1974] ECR 449). An ‘act’ need not be directly effective in order to be capable of interpretation (Mazzalai (Case 111/75) [1976] ECR 657).

The EU’s approach to interpretation

Any court in the world has a choice as to its approach to interpretation, and the ECJ is no exception. There are three main methods:

  1. Literal. The ordinary dictionary meaning. Popular with English courts, but not with the ECJ. The multilingualism situation makes this method impracticable.
  2. Contextual. Look to EU law as a whole, not just the particular piece of legislation.
  3. Purposive. Interpret the legislation in the way which most furthers the purposes of the Union. EU legislation (whether primary or secondary) lends itself to this approach because of the presence of a ‘preamble’ setting out the aims and objectives of the legislation.

Generally, the ECJ takes a ‘teleological’ approach, which may be described as a combination of the second and third approaches. The position is summed up in the following extract from Re Adidas AG (Case C–223/98) [1999] ECR I–7081:

‘Using the teleological approach allows the ECJ to update the law and meet new social and political developments.

Language differences

The Court has frequently dealt with the issue of linguistic divergences in EU legislation. For example, it held in Stauder v Ulm (Case 29/69) [1969] ECR 419:

The role of the ECJ in the preliminary reference procedure

The ECJ is supposed to be ‘reactive’, that is, it responds to questions submitted to it by the national courts. Occasionally, the ECJ will take a more ‘proactive’ approach and reformulate a question so that the answer it gives is more useful to the national court. Even more rarely, the ECJ will answer a question that was not actually asked, if the Court thinks that this will assist the national court in giving judgment. A good example of this is Marks & Spencer v Customs and Excise Commissioners (Case C–62/00) [2003] QB 866. The Court of Appeal had asked a question of the ECJ relating to Directive 77/388. The ECJ noted that the question was based on a mistaken premise regarding direct effect and, having put the Court of Appeal straight on that point, concluded that it (the ECJ) therefore needed to rephrase the question (otherwise the answer would not make sense). The Court stated:

The ECJ is not allowed to consider the validity of national law. If it is asked to do so, it may reformulate the question and return an abstract answer on the point of (EU) law involved (Costa v ENEL (Case 6/64) [1964] ECR 1141) or simply refuse to answer the question asked (see Foglia v Novello (Case 104/79) [1981] ECR 745; Bacardi-Martini v Newcastle United (Case C–318/00) [2003] ECR I–905). Nor is the Court supposed to consider how EU legislation should be applied by the national courts; however, it has done this in the past by giving ‘practical’ rulings (Stoke-on-Trent City Council v B&Q (Case C–169/91) [1993] 2 WLR 730). This is unsurprising, given that the line between interpretation and application is likely to be very fine.

The role of the national courts

The national courts, having requested a preliminary ruling, are then expected to apply it to the facts of the case and give judgment. However, legal history was made in 2003 in Arsenal FC v Reed (Case C–206/01) [2003] Ch 454, when Laddie J in the English High Court refused to apply a preliminary ruling of the ECJ that he himself had requested. Arsenal Football Club had accused Matthew Reed of infringing its trademarks by selling unofficial merchandise such as scarves and shirts bearing Arsenal’s logos (a shield and a cannon) outside the club’s ground, Highbury Stadium in north London. During the course of the subsequent trademark infringement action, Laddie J had requested a preliminary ruling on the interpretation of certain provisions in Directive 89/104. However, when the ruling was delivered, Laddie J decided that the ECJ had overstepped its interpretative jurisdiction and had made certain findings of fact (with which he disagreed) regarding the question of whether or not Arsenal supporters were likely to confuse Arsenal’s official merchandise with Reed’s unofficial merchandise. Strictly speaking, the ECJ is only supposed to make rulings on the interpretation of points of EU legislation and, thus, if Laddie J was correct then he was perfectly entitled to reach this conclusion. Nevertheless, it was a controversial decision. However, a potential crisis in the relationship between the ECJ and the High Court was averted. Having lost the case in the High Court, Arsenal FC appealed to the Court of Appeal which found, reversing Laddie J’s decision, that the ECJ had not overstepped its jurisdiction. The Court of Appeal therefore applied the preliminary ruling in full and gave judgment to Arsenal.

6.2.2 References challenging validity of EU law

Special considerations apply where the question is the validity of EU law rather than interpretation. First, the ECJ may not rule on the validity of the Treaties. It is therefore only secondary EU legislation that can be challenged on validity grounds. Second, whereas a national court may declare EU law valid and not refer, it may not declare EU law invalid (Firma Foto-Frost v Hauptzollamt Lübeck-Ost (Case 314/85) [1987] ECR 4199). The ECJ has exclusive authority in this situation. Where a national court suspects that a provision of EU secondary legislation may be invalid, therefore, a reference must be made. A good example is R (on the application of British American Tobacco) v Secretary of State for Health (Case C–491/01) [2002] ECR I–11453.

In September 2001, British American Tobacco and Imperial Tobacco sought permission from the High Court in London to apply for judicial review of ‘the intention and/or obligation’ of the UK government to transpose the Directive into national law. The application was based on several grounds, including inappropriate legislative base. The case was referred to the ECJ for a ruling, and in due course the Court held that the Directive was valid; Article 114 was the correct legislative base. Article 207 should not have been used in addition but this was a purely formal defect and did not affect the validity of the Directive.

6.2.3 ‘Docket control’: inadmissible references

The ECJ rarely refuses a request for a preliminary ruling. Provided that the question referred to the ECJ is one of interpretation, the ECJ is bound in principle to respond. However, there are three situations when requests for preliminary rulings have been declared inadmissible.

Contrived dispute

In Leur-Bloem (Case C–28/95) [1998] QB 182, the ECJ stated:

This situation arose in Foglia v Novello (No 2) (Case 244/80) [1981] ECR 3045:

This decision has been criticised, but is understandable: there was no real issue of EU law for the ECJ to determine. In Meilicke v Meyer (Case C–89/91) [1992] ECR I–4871, the ECJ, following Foglia v Novello (1981), refused to consider a series of questions referred to it from the Hanover Regional Court, as they all related to Professor Meilicke’s theories regarding EU company law and there was no genuine dispute between the parties. The Court announced that the purpose of Article 267 was to contribute to the administration of justice in the Member States, not to deliver advisory opinions on general or hypothetical questions.

In a number of subsequent cases the ECJ has accepted that the dispute was genuine, despite suggestions to the contrary by an interested observer. For example, in Idéal Tourisme (Case C–36/99) [2000] ECR I–6049 the ECJ rejected a suggestion by the Belgian government that a dispute over VAT (between a Belgian company and the Belgian tax authorities) was contrived, holding:

In Bacardi-Martini v Newcastle United FC (2003), the ECJ refused to respond to a request from the English High Court on facts not dissimilar to those in Foglia v Novello (1981).

Irrelevance

Where the request relates to provisions of EU law that are incidental to the actual dispute, the request may be refused by the ECJ. According to the ECJ in BP Supergas v Greece (Case C–62/93) [1995] ECR I–9883, a request will be refused if it is ‘quite obvious’ that the question bears ‘no relation’ to the actual subject matter of the litigation.

Insufficient information of factual/legal background

In Telemasicabruzzo (Cases C–320 to 322/90) [1993] ECR I–393, the ECJ rejected a reference that had insufficient information about the factual background or the legal dispute between the parties. This was confirmed in La Pyramide (Case C–378/93) [1994] ECR I–3999, the ECJ stating that this would be the case especially where the factual situation was complex.

6.3 The meaning of ‘court’ or ‘tribunal’

There are limitations on who can request a ruling. Only a ‘court or tribunal’ may do so. The phrase ‘court or tribunal’ has been interpreted very widely. (Article 267 has itself been interpreted by the ECJ, after national courts referred questions to it, under the Article 267 procedure!) It is certainly not required that a forum have the name ‘court’ or ‘tribunal’.

6.3.1 The Dorsch Consult factors

According to the ECJ in Dorsch Consult (Case C–54/96) [1997] ECR I–4961:

This is a very important case, as it establishes what might be described as a ‘functional’ test for establishing which bodies may invoke the Article 267 procedure. The practical result is that more bodies can seek preliminary rulings than would have been the case had the ECJ adopted a ‘literal’ approach, that is, only answering requests from bodies actually called ‘court’ or ‘tribunal’. The advantages of this ‘functional’ approach are:

  • Many bodies which do not have the name ‘court’ or ‘tribunal’ nevertheless carry out judicial functions, that is, they are deciding disputes between parties. The functional approach helps to ensure that these bodies do not have to decide on the interpretation of EU legislation themselves, which in turn means that the legislation is more likely to be applied accurately in order to resolve the disputes.
  • The functional approach means more bodies can request rulings on different provisions of EU law, which allows the ECJ to give definitive rulings on ambiguous provisions of EU legislation which may not have otherwise reached the ECJ at all.
  • The functional approach reduces the need for expensive and time-consuming appeals in the national legal systems. Often, appeals are available against decisions made by various bodies to ‘courts’. Without a functional approach, these appeals might be triggered in order to get a case into a ‘court’ in order for a request for a preliminary ruling to be made. With a functional approach, any body performing a judicial function can seek a ruling itself.

A good example of the ‘functional’ approach is Broekmeulen (Case 246/80) [1981] ECR 2311. Dr Broekmeulen’s registration as a GP had been refused. His appeal, to the Appeals Committee of the Royal Netherlands Society for the Protection of Medicine, was based on EU law. References were made to the ECJ, one of which asked whether or not the Appeals Committee was a ‘court or tribunal’. The ECJ held:

The wide scope of ‘court or tribunal’ can be seen in the following cases:

  • Royal Copenhagen (Case C–400/93) [1995] ECR I–1275 – administration board in Copenhagen, Denmark;
  • O’Flynn v Adjudication Officer (Case C–237/94) [1996] ECR I–2617 – Social Security Commissioner, UK;
  • Gebhard (Case C–55/94) [1995] ECR I–4165 – Milan Bar Council;
  • El-Yassini (Case C–416/96) [1999] ECR I–1209 – immigration adjudicator, UK;
  • Abrahamsson and Andersson (Case C–407/98) [2000] ECR I–5539 – Universities Appeals Board, Sweden;
  • Cadbury Schweppes (Case C–196/04) [2006] ECR I–7995 – Special Commissioners of Income Tax, UK;
  • Jia (Case C–1/05) [2007] ECR I–1 – Alien Appeals Board, Sweden;
  • Torresi (Cases C–58 and 59/13) [2015] QB 331; [2015] 2 WLR 29 – National Bar Council, Italy.

In El-Yassini (1999), the ECJ gave very careful consideration to the question of whether or not an immigration adjudicator in the UK qualified as a ‘court or tribunal’. In the end it decided that an adjudicator did qualify. Read this quote from the judgment and note how many of the Dorsch Consult (1997) factors were satisfied:

Conversely, in Nordsee (Case 102/81) [1982] ECR 1095, an independent arbitrator was held not to be a court. This was because the arbitrator lacked compulsory jurisdiction. A similar decision was reached in Denuit & Cordenier (Case C–125/04) [2005] ECR I–923, where the Court stated:

In Procura Della Republica v X (Case C–74/95) [1996] ECR I–6609, the ECJ held that questions referred to it by the Italian Public Prosecutor were inadmissible, as he did not constitute a ‘court or tribunal’. And in Victoria Film A/S (Case C–134/97) [1998] ECR I–7023 a reference from a body within the Swedish tax administration was held inadmissible (it did not carry out a judicial function). Rather, the body carried out a purely administrative function. If anything, the body tried to prevent disputes arising in the first place, as opposed to resolving disputes which had already arisen.

Even a court may not be a ‘court’ if it is carrying out an administrative (as opposed to judicial) function. In Salzmann (Case C–78/99) [2001] ECR I–4421, the ECJ held that the District Court, Bregenz, Austria, was not a ‘court’ when it was acting as a land registry. Similarly, in Lutz & Others (Case C–182/00) [2002] ECR I–547, the Regional Court, Wels, Austria, was not a ‘court’ when acting as a company’s registry. The ECJ held:

6.3.2 ‘Independence’

This criterion has generated some important case law. In some Member States, a number of tribunals are closely connected with government departments whose decisions they are called upon to examine. Does this satisfy the criterion of independence? This issue arose in a case involving the Austrian legal system. In Köllensperger & Atswanger (Case C–103/97) [1999] ECR I–551, a reference had been made by the Procurement Office of the Land of Tyrol, Austria. Its members were appointed by the Tyrol government, and could be removed ‘if the conditions for appointment are no longer met or if circumstances occur which prevent proper exercise of the office and are likely to do so for a long time’. According to the ECJ, this ‘appears prima facie too vague to guarantee against undue intervention or pressure on the part of the executive’. However, the Court found that there were guarantees of independence in other provisions of Austrian law, including a provision expressly prohibiting the giving of instructions to members of the Procurement Office in the performance of their duties.

Two other cases illustrate the problem of ensuring ‘independence’. In the first case, Gabalfrisa & Others (2000), the request was declared admissible, but in the second case, Schmid (2002), the Court declared that the ruling was inadmissible.

In Schmid, the ECJ held that a body cannot be regarded as an independent ‘court or tribunal’ for the purposes of Article 267 where it has ‘an organisational and functional link’ with a government department whose decisions it is called upon to review.

In Wilson (Case C–506/04) [2006] ECR I–8613, the Court offered extensive guidance on the ‘concept of independence’, as follows:

One of the key criteria for judicial independence identified in Wilson is protection ‘against external intervention or pressure’. It was the lack of that specific criterion that led the Court to refuse to answer questions submitted to it by the Spanish Central Tax Tribunal (CTT) in Proceedings Brought by Banco de Santander (Case C–274/14) [2020] 2 CMLR 29. The CTT’s members were appointed by Royal Decree, on the proposal of the Spanish Minister for the Economy and Finance, for an indefinite period. Crucially, however, they could also be removed by exactly the same method. Hence, there were no ‘express legislative provisions, such as those applicable to members of the judiciary’ in Spain generally, governing the removal of CTT members. The members were therefore not ‘protected against direct or indirect external pressures that are liable to cast doubt on their independence’. The CTT was therefore not a ‘court or tribunal’ and its questions were inadmissible.

The importance of judicial independence was recently re-stated in Associação Sindical dos Juízes Portugueses (Case C–64/16) [2018] 3 CMLR 16, in which the Court stated:

6.4 The discretionary reference procedure

The decision on when to refer questions, and on what issues, is left entirely up to the national courts (Pigs Marketing Board v Redmond (Case 83/78) [1978] ECR 2347). It is normal, but not essential, that one or more of the parties will have attempted to rely upon some provision of EU legislation during the case. However, the national court may issue a reference of its own volition if it deems it necessary to reach a decision (Verholen (Cases C–87 to 89/90) [1991] ECR I–3757). It is essential that the request for a ruling be made while the case is still proceeding in the national court. After that point it is too late, because the ECJ decision would no longer be ‘necessary’ to enable the national court to give judgment (Pardini (Case 338/85) [1988] ECR 2041).

For those national courts or tribunals falling within the second paragraph of Article 267 (and this is the vast majority of them), there is a discretion whether or not to refer the case to the ECJ (note the word ‘may’). This discretion cannot be removed by national rules as to precedent (Rheinmuhlen-Dusseldorf (Case 166/73) [1974] ECR 33). A lower court may refer a matter to the ECJ despite a superior court’s ruling to the contrary.

The fact that the ECJ has already decided a particular matter should not of itself prevent a further reference. In Da Costa en Schaake (Cases 28 to 30/62) [1963] ECR 61, the ECJ declared it had the right to depart from previous decisions.

A recent case has demonstrated the absolute, fundamental importance of the national court not having its discretion fettered. In Commission v Poland (Case C–791/19) [2022] 1 CMLR 15, the European Commission alleged that Polish legislation which created the possibility that Polish judges seeking preliminary rulings could be made subject to disciplinary proceedings was contrary to Article 267(2). The Court of Justice agreed. The Court stated:

6.5 The mandatory reference procedure

6.5.1 Introduction

Under the second paragraph of Article 267, the court or tribunal ‘may’ make a request; under the third paragraph of Article 267, courts or tribunals, against whose decisions there is no judicial remedy under national law, ‘shall’ refer. The importance of Article 267(3) was explained by the ECJ in Commission v France (Case C–416/17) (2018) (unreported) in the following terms: ‘The obligation to make a reference laid down in [Article 267(3)] is intended to prevent a body of national case-law that is not in accordance with the rules of EU law from being established in any of the Member States’.

It is crucial to note that only for certain courts or tribunals is it mandatory to refer. Such courts may be referred to using the shorthand ‘courts of last resort’. The question is: which courts are they? In Costa v ENEL (Case 6/64) [1964] ECR 1141, a request had come from an Italian magistrates’ court. There was no appeal from the magistrates’ decision, because of the small amount of money involved. The ECJ stated (emphasis added): ‘By the terms of this Article … national courts against whose decision, as in the present case, there is no judicial remedy, must refer the matter to the Court of Justice.’

Although slightly ambiguous, this has been taken to imply that certain courts, although generally subject to the second paragraph, could find themselves subject to the third paragraph if no appeal was available in a particular case. This issue has been examined by the ECJ, in the context of the Swedish judicial system.

The ECJ held that these procedural rules did not convert the Swedish Court of Appeal into a ‘court of last resort’:

The decision in Lyckeskog was followed and applied in Cartesio (Case C–210/06) [2008] ECR I–9641. That case raised the question whether the Regional Court of Appeal in Hungary was subject to the third paragraph of Article 267, given that its decisions were final, subject to an ‘extraordinary’ appeal to the Hungarian Supreme Court. The ECJ ruled that, because appeals were available, albeit only in limited circumstances, the Regional Court of Appeal was not subject to the mandatory reference procedure.

6.5.2 Mandatory references and hypothetical questions

Although national supreme courts are obliged to make references to the ECJ when a question is raised before them, that does not mean they have to seek rulings if the question raised is actually irrelevant to the case. However interesting the question may be, if it is not essential to the outcome of the case it is hypothetical. In CILFIT (Case 283/81) [1982] ECR 3415, the ECJ held:

6.5.3 Mandatory references and previous rulings

Does a national court of last resort have to make a preliminary reference even if the point has already been decided? If so, that would generate a large number of ‘repeat’ references, which would be very inefficient. Hence, the ECJ has decided that national courts of last resort do not have to make references in such cases. In Da Costa (1963), the ECJ stated:

This was subsequently endorsed by the ECJ in CILFIT (1982). The ECJ referred to Da Costa and added that ‘the same effect … may be produced … even though the questions at issue are not strictly identical’. And in Bulmer v Bollinger (1974), Lord Denning said:

A good example of this was seen in the pre-Brexit UK Supreme Court case of Mirga; Samin [2016] UKSC 1; [2016] 1 WLR 481. Despite being the UK’s ‘court of last resort’, the Supreme Court declined to seek a preliminary ruling because the question raised had already been decided by the ECJ in Dano (Case C–333/13) [2015] 1 CMLR 48; [2015] 1 WLR 2519 and Alimanovic (Case C–67/14) [2016] 1 CMLR 29; [2016] QB 308. All of these cases will be discussed later (see Chapter 11).

6.5.4 Mandatory references and acte clair

According to the ECJ in CILFIT (1982), ‘the correct application of [Union] law may be so obvious as to leave no scope for any reasonable doubt as to the manner in which the question raised is to be resolved’. This would entitle a national ‘court of last resort’ to decide not to invoke the Article 267 procedure. A decision not to request a ruling because the provision is ‘so obvious as to leave no scope for reasonable doubt’ is an example of acte clair (literally, ‘clear act’), a doctrine developed from French law. However, in CILFIT (1982), the ECJ gave very clear instructions that acte clair must be used with caution:

The Court went on to list those conditions in the next paragraph:

The instruction not to abuse acte clair was subsequently emphasised in Inter-modal Transports (Case C–495/03) [2005] ECR I–8151, where the Court stated that the national court or tribunal must ‘in particular’ be convinced that the other Member States’ national courts and the ECJ itself would find the matter ‘equally obvious’.

In X & Van Dijk (Cases C–72, 197/14) [2016] 1 CMLR 27, the Supreme Court of the Netherlands asked the ECJ whether it was entitled to treat the interpretation of a provision of EU legislation as acte clair (which it was inclined to do) notwithstanding the fact that a lower court had, coincidentally, requested a preliminary ruling on the same provision (albeit in a different case) and therefore did not regard the matter as acte clair. The ECJ held that CILFIT gave the national court of last resort ‘sole responsibility’ for determining whether the correct application of EU law was so obvious as to leave no scope for any reasonable doubt. It therefore followed that it was for the national court of last resort ‘alone’ to take upon themselves ‘independently’ the responsibility for determining whether the provision of EU legislation in question was acte clair. The fact that a lower court had requested a ruling did ‘not preclude the supreme court of a Member State from concluding, from its examination of the case and in keeping with the criteria laid down in the judgment in CILFIT, that the case before it involves an acte clair’.

In Ferreira da Silva e Brito & Others v Portugal (Case C–160/14) [2016] 1 CMLR 26, decided on the same day as X & Van Dijk, the ECJ reiterated that, in general, ‘the fact that other national courts or tribunals have given contradictory decisions is not a conclusive factor capable of triggering the obligation’ set out in Article 267(3). The Court added that a national court of last resort ‘may take the view that, although the lower courts have interpreted a provision of EU law in a particular way, the interpretation that it proposes to give of that provision, which is different from the interpretation espoused by the lower courts, is so obvious that there is no reasonable doubt’. However, in the case in hand, the ECJ held that the Portuguese Supreme Court should have referred the question, involving the interpretation of the phrase ‘transfer of an undertaking’ in Directive 2001/23, to the ECJ under Article 267(3). This was because of the ‘conflicting lines of case law at national level’ combined with ‘the fact that that concept frequently gives rise to difficulties of interpretation in the various Member States’. (This aspect of the case is discussed in more detail in Chapter 9 on State Liability.)

More recently, in Consorzio Italian Management v Rete Ferroviaria Italiana (No.2) (Case C–561/19) [2022] 2 CMLR 6, the ECJ held that that the ‘mere fact’ that a provision of EU law may be interpreted in ‘another way or several other ways’ did not necessarily create a reasonable doubt as to the correct interpretation of that provision. Nonetheless, where the national court of last resort was made aware of ‘the existence of diverging lines of case-law – among the courts of a Member State or between the courts of different Member States – concerning the interpretation of a provision of EU law’, that court had to be ‘particularly vigilant’ in its assessment of whether or not there was any reasonable doubt as to the correct interpretation of that provision. In the same case, the Court also signalled a slight softening of one of the CILFIT criteria when it said:

While a national court or tribunal of last instance cannot be required to examine each of the language versions of the provision in question, the fact remains that it must bear in mind those divergences between the various language versions of that provision of which it is aware, in particular when those divergences are set out by the parties and are verified. [emphasis added]

Notwithstanding these instructions from the ECJ to national courts of last resort not to abuse acte clair, there are many examples of national courts doing exactly that. The UK Supreme Court, and its predecessor the House of Lords, were notorious for taking liberties with CILFIT. However, the UKSC was not alone here. For example, Turmo has criticised the Conseil d’É tat in France for abusing acte clair. He wrote that:

An excessively generous interpretation of the acte clair doctrine remains a problematic element in the Conseil d’É tat’s case law (and that of other national supreme courts)…. The Conseil d’É tat and other national supreme courts would do well to revise their interpretation of [acte clair] and exercise greater restraint in their own engagement with the Court’s case law.

Araceli Turmo, ‘A Dialogue of Unequals – the European Court of Justice reasserts National Courts’ Obligations under Article 267(3) TFEU’ (2019) 15 ECL Review 340

6.5.5 Need for national ‘courts of last resort’ to explain a failure to refer

On several occasions, the European Court of Human Rights in Strasbourg has held that if a national ‘court of last resort’ fails to seek a preliminary ruling from the ECJ in Luxembourg, whether on the grounds of irrelevance, that the question had already been answered, or on the basis of acte clair, but fails to explain why it has failed to refer, this may amount to a breach of the right to a fair trial in Article 6 of the European Convention of Human Rights (see for example Ullens de Schooten & Rezabek v Belgium (2011), Dhahbi v Italy (2014) and Schipani v Italy (2015)). Broberg has summarised the situation:

Where an EU Member State court refuses to submit a preliminary reference to the CJEU, an infringement of the right to a fair trial provided for in art.6(1) of the ECHR may be found. When an EU Member State court of last instance arbitrarily refuses to submit a preliminary reference, as a main rule, this will infringe art.6(1) of the ECHR.

Morten Broberg, ‘National EU courts must seek advice in Luxembourg or face reproach in Strasbourg’ [2021] 2 EHRLR 162

6.5.6 Academic reaction to acte clair

Academic reaction to the doctrine of acte clair itself has been mixed. The consensus seems to be that the ECJ in CILFIT (1982) was right to endorse acte clair. However, opinion is divided about the criteria to be satisfied by national courts before invoking the doctrine.

The effect of the CILFIT decision … would be to enable national judges to justify any reluctance they might feel to ask for a preliminary ruling…. Of the factors to be borne in mind by national courts before they concluded that the meaning of a provision of [Union] law was clear, only the requirement that the different language versions be compared … had any teeth…. In short, the overall effect of CILFIT would be to encourage national courts to decide points of [Union] law for themselves.

A Arnull, ‘The Use & Abuse of Article 177 EEC’ (1989) 52 MLR 622

On the other hand, Professor Rasmussen was perhaps the most highly critical commentator of CILFIT (1982). While he endorsed the ECJ’s decision to allow national courts of last resort to decide questions of EU law for themselves, he believed that the CILFIT (1982) criteria were so stringent that, taken as a whole, the judgment achieved the opposite of what the Court said it intended to do. Writing in 1984, he said:

The author believes that … CILFIT means something very different from what it prima facie establishes…. The real strategy is different from the apparent strategy. The real strategy of CILFIT is not to incorporate an acte clair concept into [Union] law. It is to call the national judiciaries to circumspection when they are faced with problems of interpretation and application of [Union] law.

H Rasmussen, ‘The European Court’s Acte Clair Strategy in CILFIT’ (1984) 9 ELR 242

He repeated his criticism in 2000, calling for the ECJ to rewrite the CILFIT (1982) judgment to make acte clair more easily available:

The Court of Justice’s present predicaments are self-inflicted … I refer to the submission straightjacket designed by the Court in CILFIT. This judgment has functioned as a magnet, drawing numerous, and often less-than-necessary, cases up to the Court. The ECJ ought … to rewrite CILFIT’s submission criteria, thereby watering down some of the stringency of the conditions to which it subjects national courts’ duty to make use of [Article 267].

H Rasmussen, ‘Remedying the Crumbling EC Judicial System’ (2000) 37 CMLR 1071

6.5.7 No other exceptions to mandatory referral

In Consorzio Italian Management v Rete Ferroviaria Italiana (No.2) (Case C–561/19) (2022), the ECJ confirmed that the only situations in which a national court of last resort could avoid mandatory referral were those set out in Da Costa and CILFIT above. There were no other exceptions. It therefore followed that, simply because a preliminary ruling had already been requested from and answered by the ECJ in a particular case, it did not relieve the national court of seeking another ruling in the same case if a question concerning the interpretation of EU law, the answer to which was necessary for the resolution of the dispute, remained after the Court’s decision.

6.6 Reform of the preliminary reference procedure Reform of reference procedure

The ECJ is faced with a very large backlog of cases, such that the average waiting time (from a national court or tribunal requesting a preliminary ruling until the ruling itself) is approximately 15 months. Although the average waiting time has reduced from an all-time high of 25 months in 2003 (largely thanks to the appointment of new judges following the 2004 and 2007 accessions), the fact that the typical case involves a wait of more than a year means that national judges may be inhibited from asking questions, instead attempting to answer the questions themselves, which threatens both individual rights (if the judges get the answer wrong) and the ‘co-operation’ between the various courts.

One thing about which there is academic agreement is that the preliminary reference procedure is in need of reform. According to Professor Rasmussen (‘Remedying the Crumbling EC Judicial System’ (2000) 37 CMLR 1071): ‘It is a generally shared view today that the case for a comprehensive and profound judicial reform has become compelling.’ Johnston (‘Judicial Reform and the Treaty of Nice’ (2001) 38 CMLR 499) agrees: ‘Clearly, there is a serious workload problem for the Court, due to a number of factors. It is particularly serious in the context of references for a preliminary ruling.’

The delay has been caused by, among other things:

  • The expansion of the EU (from six in 1958 to nine in 1973, to ten in 1981 and 12 in 1986, to 15 in 1995 and 25 in 2004, to 27 in 2007, and 28 Member States in 2013), allowing more courts and tribunals to refer questions. (Of course, since Brexit, that number has dropped back to 27.)
  • The associated growth in the number of official EU languages (from four in 1958 to six in 1973, to seven in 1981, nine in 1986, 11 in 1995, and now 24). This imposes massive burdens on the ECJ in terms of the translation of judgments.
  • The width of the ECJ’s own definition of ‘court or tribunal’ in cases such as Dorsch Consult.
  • The ever-increasing scope and volume of EU secondary legislation. For example, during 2003 two new discrimination Directives (Directive 2000/43, the ‘Race Directive’, and Directive 2000/78, the ‘Framework Directive’) entered into force. Both Directives – especially the Framework Directive – have generated a number of preliminary rulings over the years and are likely to generate many more in the years to come. These Directives will be examined in detail in Chapter 18.
  • The ECJ’s development of concepts such as fundamental rights and state liability, and the reliance placed on the Charter of Fundamental Rights (2000), all of which have triggered waves of new requests for preliminary rulings.

Against this, the ECJ’s own tentative steps to reduce its workload through acte clair in CILFIT in 1982 has had negligible impact. All of this has led to a net increase in the number of preliminary references which the Court has to deal with, as the table shows:

Year Number of preliminary references
1961 1
1971 37
1981 108
1991 186
2001 237
2011 423
2019 641 (an all-time high)

6.6.1 Conferral of jurisdiction on the General Court

In the future, some preliminary rulings may be transferred to the General Court, albeit in ‘specific areas’ and with the possibility of reference to/review by the ECJ. The procedure is set out in Article 256(3):

Although the General Court has been in operation since 1989, it has never actually handled preliminary rulings. When the General Court was created, it had no jurisdiction in this area. That changed when Article 256(3) was inserted into what was then the EC Treaty in 2003. However, the procedure has not yet been brought into operation, the main stumbling block being the lack of any decision on what the ‘specific areas’ should be. Numerous suggestions have been made (see in particular P Dyrberg, ‘What Should the Court of Justice be Doing?’ (2001) 26 EL Rev 291) but, as yet, none of the suggestions has been adopted.

Assuming that the General Court is allowed to start hearing preliminary rulings at some point, there are likely to be teething problems until the new procedure is fully developed. For example:

  • The General Court ‘may’ refer cases on to the ECJ when a ‘decision of principle’ is involved – but what does that mean?
  • Although the ECJ is only to review General Court rulings ‘exceptionally’, this nevertheless undermines the authority of all those rulings. Can any national court apply any General Court ruling whilst a possible review is still pending? Should the ECJ be given a time limit by which it must decide whether it is to invoke its review power?

Academic reaction to the conferral of preliminary ruling jurisdiction on the General Court has been mixed, although mostly positive. Tridimas described the procedure as setting an ‘acceptable balance between competing demands’ – the need for consistent interpretation of EU legislation on one hand, and the need for a reduction in the waiting time on the other. He described the idea of the ECJ and General Court sharing jurisdiction over preliminary rulings as ‘preferable over alternative reforms’ (T Tridimas, ‘Knocking on Heaven’s Door: Fragmentation, Efficiency and Defiance in the Preliminary Reference Procedure’ (2003) 450 CML Rev 9). Another commentator described the new procedure as ‘very important’, adding ‘the sooner this possibility is exercised the better’ (B Vesterdorf, ‘The Community Court System Ten Years from Now and Beyond: Challenges and Possibilities’ (2003) 28 EL Rev 303). However, Heffernan was less enthusiastic, suggesting that ‘there is every reason to believe that [the General Court’s] contribution will be limited’ (L Heffernan, ‘The Community Courts Post-Nice: A European Certiorari Revisited’ (2003) 52 ICLQ 907).

6.6.2 The urgent preliminary ruling procedure

In March 2008, a new urgent preliminary rulings procedure was adopted, to be used in exceptional cases. The Court of Justice explained the thinking behind the new procedure as follows:

This procedure is applicable as from 1st March 2008 and should enable the Court to deal far more quickly with the most sensitive issues relating to the area of freedom, security and justice, such as those which may arise, for example, in certain situations where a person is deprived of his liberty and the answer to the question raised is decisive as to the assessment of the legal situation of the person detained or deprived of his liberty; or, in proceedings concerning parental authority or custody of children, where the jurisdiction under [Union] law of the court hearing the case depends on the answer to the question referred for a preliminary ruling.

The new procedure cuts the waiting time by:

  • restricting the number of parties entitled to submit written observations;
  • referring all cases involving the area of freedom, security and justice to a special chamber of five judges who will decide whether to apply the new procedure and, if they decide to do so, to give their ruling shortly afterwards;
  • using only electronic communication.

The new procedure was used for the first time in Rinau (Case C–195/08 PPU) [2008] ECR I–5271; [2009] 2 WLR 972. On 30 April 2008, a preliminary ruling was requested by the Supreme Court of Lithuania in a case involving child custody. Luisa Rinau’s parents were divorced and she lived with her mother in Lithuania, but her father, who lived in Germany, was seeking custody. The ECJ applied the new procedure and the ruling was delivered on 11 July, a little over ten weeks later. The Court explained that the urgency was justified by the need to protect Luisa from harm, the need to avoid damaging her relationship with her father, and the need to ensure a ‘fair balance’ between Luisa’s interests and those of her parents.

An even faster turnaround was achieved in Santesteban Goicoechea (Case C–296/08 PPU) [2008] ECR I–6307. Here, the Court of Appeal in Montpellier, France, requested a preliminary ruling in a case involving the possible extradition of a Spanish national from France to face criminal charges in Spain. The ruling was requested on 3 July 2008 and the ECJ ruling was given on 12 August, less than six weeks later. Here, the urgency was justified because the individual concerned was being held in detention in France following the completion of a prison sentence there pending his possible extradition to Spain.

Academic reaction to the new procedure has been positive, but Koutrakos (‘Speeding up the Preliminary Reference Procedure: Fast but not too Fast’ (2008) 33 EL Rev 617) has pointed out that ‘there is a balance which must be struck’ between delivering judgments as quickly as possible (on the one hand) and providing sufficient time for the ECJ’s judges ‘to reflect on the questions put before them, assess the arguments … and consider the wider ramifications of their conclusions’ (on the other). He notes that taking time to deliver a judgment is not a ‘luxury’ but ‘an essential prerequisite for the proper administration of justice’.

6.6.3 Other reform proposals

Numerous proposals for reform of the preliminary rulings procedure have been advanced over the years. These proposals fall into two distinct groups:

  1. those proposals which seek to reduce the volume of requests coming into the ECJ; and
  2. those that seek to boost or streamline the European Union’s judicial capacity for dealing with these cases.

(The conferral of jurisdiction on the General Court, noted above, falls into the latter group.) Group 1 proposals include:

  • Restricting the range of national courts or tribunals with the discretion to seek rulings. This could entail removing the right of ‘first instance’ national courts and tribunals (such as magistrates’ and county courts plus most tribunals in England) from seeking rulings. This would probably have little practical impact as ‘first instance’ courts tend not to seek many references anyway.
  • Removing the right of ‘first tier’ appeal courts and tribunals from seeking rulings. This would probably have a significant impact in terms of cutting down requests. Statistics indicate that a large proportion of the preliminary rulings come from such bodies. However, if these bodies lost the right to seek rulings, they would have to interpret EU legislation themselves (with the risk of divergent rulings being made).
  • Tightening up the Dorsch Consult (1997) criteria, thereby cutting out requests from bodies such as ‘appeal boards’ and ‘adjudicators’.
  • Abolishing mandatory referral for national courts of last resort. Arguably, this has already been achieved through the doctrine of acte clair.
  • Rewording Article 267 to require national judges to consider the importance, difficulty and/or novelty of the ruling requested.

Group 2 proposals include:

  • Allowing the ECJ to filter questions. Most national courts of last resort can filter out cases deemed to be insufficiently important or novel. There are grounds for saying that the ECJ should be given the same power.

  • Setting up regionalised courts with EU legal specialism, with the ECJ restyled as a ‘European High Court of Justice’ overseeing the new regionalised courts. This is not dissimilar to the organisation of federal courts in the United States, with the Supreme Court in Washington DC at the apex, overseeing the decisions of the various federal ‘circuit’ courts, which in turn oversee the federal courts in the states within each ‘circuit’. The latter suggestion was first proposed by the academics Jacque and Weiler:

    At the apex of the system will remain the [ECJ], renamed perhaps as the European High Court of Justice. We propose the creation of four new Community Regional Courts which will have jurisdiction to receive preliminary references from, and issue preliminary rulings to, national courts within each region. Upon the decision or preliminary ruling of the Regional Court being issued, a party to the proceedings and the Commission, Council, Parliament or Member States as interveners, may appeal to the European High Court of Justice.

    J P Jacque and J Weiler, ‘On the Road to EuropeanUnion: A New Judicial Architecture’ (1990) 27 CMLR 185

The academics Veerle Heyvaert, Justine Thornton and Richard Drabble proposed a different version of decentralisation. Extrapolating from the creation of the General Court in 1989 and the Civil Service Tribunal in 2005 (subsequently abolished), they argued that:

It would be possible to expand the range further by establishing specialised bodies for questions that are typically considered to require a high level of technical specialisation and expertise to resolve, such as trademarks and competition disputes. Environmental decision-making might qualify as a domain for which a specialised court is needed. The addition of new EU courts or tribunals could further cut down on delays or, accounting for the continuing rise in proceedings lodged over time, at least keep them on an even keel.

V Heyvaert, J Thornton and R Drabble, ‘With Reference to the Environment: The Preliminary Reference Procedure, Environmental Decisions and the Domestic Judiciary’ (2014) 130 LQR 413

All of these proposals would reduce the workload of the ECJ and hence reduce the time delay, but at what cost?

  • Restricting national courts’ ability to seek rulings will probably lead to more appeals in the national courts as disappointed litigants try to get a case into a national court which has retained the power to request rulings.
  • Filtration may undermine the ‘co-operation’ that exists between the ECJ and the Member States’ courts and tribunals. This may deter national courts from seeking rulings in cases where, now, they would seek such a ruling, possibly leading to inconsistency of interpretation in different Member States and stifling the development of EU legal principles (key doctrines such as direct effect, indirect effect, state liability and fundamental rights were all developed during preliminary ruling cases).
  • Regionalisation threatens uniformity, and setting up several new courts will incur considerable cost in terms of infrastructure, staffing, communications and IT. There may also be squabbles as to the location of any new regional courts.

SUMMARY

  • Article 267 TFEU allows any ‘court or tribunal’ of a Member State to request the Court of Justice to interpret provisions of EU legislation. The national court or tribunal then has the task of applying that law, as interpreted. Having one court to interpret all EU legislation ensures uniform application (Stauder v Ulm).
  • The Court usually adopts a ‘purposive’ approach to interpretation, seeking to interpret ambiguous EU law in a way which most promotes the purpose behind the legislation (Adidas).
  • Only a ‘court or tribunal’ may request a ruling, but this means that the body must be carrying out a judicial function, taking into account various factors such as permanence, an inter partes procedure, and compulsory jurisdiction (Dorsch Consult). Independence is crucial (Schmid).
  • Arbitrators, public prosecutors and administrative bodies are not a ‘court or tribunal’ (Nordsee, Victoria Film). Even a court may not be a ‘court’ if it is carrying out an administrative function (Salzmann, Lutz).
  • Under Article 267(2), any court or tribunal ‘may’ request a ruling. This gives national courts and tribunals an absolute discretion, which cannot be fettered by national law (Commission v Poland). Under Article 267(3), ‘courts of last resort’ must do so where a point of EU law is involved, subject to three exceptions: previous rulings (Da Costa), irrelevance (CILFIT) and ‘acte clair’, i.e. the national court is convinced that the answer to the question is obvious (CILFIT). Acte clair must be used with caution. In particular, the national court must be convinced that the matter is equally obvious to the courts of the other Member States and to the Court of Justice (Intermodal Transports). There are no other exceptions to Article 267(3) (Consorzio Italian Management (No.2)).
  • The ECJ is bound in principle to respond to a request for a ruling, subject to requests declared to be inadmissible: contrived disputes (Foglia v Novello), irrelevance, lack of background factual and/or legal information.
  • Under the urgent procedure, certain cases may be dealt with very quickly. Typical cases involve child custody (Rinau) or where a person is in custody (Santesteban Goicoechea).
  • The ECJ is faced with a very large backlog of cases; the average waiting time (from a national court requesting a ruling until the ruling itself) is around 15–16 months. If the delay is not reduced further, national judges may be inhibited from asking questions, instead attempting to answer the questions themselves, which threatens the whole point of the procedure and the ‘co-operation’ between the ECJ and the national courts.
  • The delay has been caused by, inter alia, the expansion of the EU (from 6 to 28 Member States, although that has dropped back to 27 following Brexit), allowing more courts and tribunals to refer questions; the associated growth in the number of official EU languages; the width of the ECJ’s own functional definition of ‘court or tribunal’; and the ever-increasing scope and volume of EU secondary legislation.
  • One solution is to confer jurisdiction on the General Court in ‘specific areas’ (Article 256(3) TFEU). As yet, no ‘specific areas’ have been decided. Under this procedure, there will be the possibility of cases involving ‘decisions of principle’ being referred on to/reviewed by the ECJ.
  • Other proposals for reform include appointing more judges; restricting the range of national courts and tribunals with the discretion to seek rulings; abolishing mandatory referral for ‘courts of last resort’; filtering questions; and setting up regional courts with EU legal specialism (decentralisation). All of these proposals would reduce the workload of the ECJ and hence reduce the time delay, but there are risks, especially to consistency.

Further reading

Articles

  • Anagnostaras, G, ‘Preliminary Problems and Jurisdiction Uncertainties: The Admissibility of Questions Referred by Bodies Performing Quasi-judicial Functions’ (2005) 30 EL Rev 878.
  • Arnull, A, ‘Judicial Architecture or Judicial Folly? The Challenge Facing the European Union’ (1999) 24 EL Rev 516.
  • Barnard, C, ‘The PPU: Is it Worth the Candle? An Early Assessment’ (2009) 34 EL Rev 281.
  • Broberg, M and Fenger, N, ‘If You Love Somebody Set them Free: On the Court of Justice’s Revision of the Acte Clair Doctrine’ (2022) 59 CML Rev 711.
  • Butler, G, ‘Independence of Non-judicial Bodies and Orders for a Preliminary Reference to the Court of Justice’ (2020) 45 EL Rev 870.
  • Jacque, JP and Weiler, J, ‘On the Road to European Union: A New Judicial Architecture’ (1990) 27 CMLR 185.
  • Johnston, A, ‘Judicial Reform and the Treaty of Nice’ (2001) 38 CMLR 499.
  • Komarek, J, ‘In the Court(s) We Trust? On the Need for Hierarchy and Differentiation in the Preliminary Ruling Procedure’ (2007) 32 EL Rev 467.
  • Lord Mance, ‘The Interface between National and European Law’ (2013) 38 EL Rev 437.
  • Rasmussen, H, ‘Remedying the Crumbling EC Judicial System’ (2000) 37 CMLR 1071.
  • Reyns, C, ‘Saving Judicial Independence: A Threat to the Preliminary Ruling Mechanism?’ (2021) 17 ECL Rev 26.
  • Vesterdorf, B, ‘The Community Court System Ten Years from Now and Beyond: Challenges and Possibilities’ (2003) 28 EL Rev 303.

7 The relationship between EU law and national law – supremacy

DOI: 10.4324/9781003218562-7

7.1 The origins of supremacy and the link with supranationalism

7.1.1 The basic meaning of ‘supremacy’

The EU is a unique international organisation based on a co-operative partnership between its 27 Member States. The latter voluntarily pooled their resources and sovereign powers in certain areas of policy (e.g. agriculture, fisheries, competition, etc.) and have tasked the supranational institutions of the EU (e.g. the Council, the European Parliament, the Commission, the ECJ, etc.) to drive its relevant policy agendas by legislating and regulating the activities of the Union. The various fields of competence of the EU represent areas of mutual and common interests which the Member States have decided are better addressed collectively via co-operation. These fields are divided into three main categories: Article 3 TFEU exclusive competencies (e.g. customs union, monetary policy, common commercial policy); Article 4 TFEU shared competences (e.g. internal market, social policy, energy, environment); and Article 6 TFEU supporting competencies (e.g. health, culture, tourism, education). In these areas of transferred powers given to the EU institutions by the Member States, the latter have thus agreed to limit their ability to legislate unilaterally.

On that basis, ‘supremacy’ or ‘primacy’ means that, in areas where EU law is relevant to a case before a national court (i.e. in a matter which is within the sphere of competence of the EU), EU law prevails over national law. In that sense, primacy of EU law is a question of jurisdiction – in matters pertaining EU law jurisdiction, EU law shall apply; similarly in matter of national jurisdiction only, national shall prevail. Supremacy and primacy are terms that have been used interchangeably; although the term ‘supremacy’ has not (or incredibly rarely) been used in the jurisprudence of the ECJ. Therefore, it is argued that primacy is the more accurate terminology.

The doctrine of primacy has been said to mirror, in some senses, the way that parliamentary law in England and Wales, in the form of Acts of Parliament, prevails over all other forms of English law. The practical consequence, then, of this supremacy/primacy is that wherever there is a conflict between the national law of a Member State and EU law itself, it is EU law that must be applied.

Supremacy/primacy is one of the key doctrines, or constitutional pillars, created by the European Court of Justice that ensure the enforceability and efficacy of the Union legal order, which is based on the international agreement known as the Founding Treaties. The doctrine or primacy is further supported by the doctrine of direct effect, which also encompasses the principles of indirect effect and state liability.

It would be fair to describe all of these as fundamental principles without which EU law could not have developed. So the European Court of Justice has been active in ensuring the success of the EU and its gradual evolution.

Supremacy/primacy has probably become one of the most entrenched of all principles associated with EU law. Nevertheless, despite the inevitable political tensions between Member States and the EU institutions, and some testing times between the Court of Justice and the national courts, it is possibly also fair to say that it has ultimately proved to be one of the most enduring and engaging doctrine.

7.1.2 The reasons for a doctrine of supremacy

While there is no actual mention of the principle within the Treaties, the doctrine has been recognised in Declaration 17 concerning primacy, annexed to the Treaties; and it is simple to see why the European Court of Justice created and then developed the principle of supremacy/primacy:

  • The fundamental objective of creating a single Internal Market demands that there should be harmonisation between Member States, which in itself then depends on a uniform application of EU law within the Member States.
  • The whole structure of the Community was founded on the idea of supranationalism. Without supremacy/primacy of EU law, the institutions would be deprived of supranational effect; EU law uniformity and efficacy would be sacrificed to national self-interest.

In this way the real justifications for the existence of a doctrine of supremacy are twofold:

  • First, it prevents any possibility of a questioning of the validity of EU law within the Member States themselves to ensure the effectiveness of the European edifice.
  • Second, it fulfils what is sometimes referred to as the ‘doctrine of pre-emption’, and it does so in two ways:
    • the fact that EU law is supreme means that the national courts in Member States are prevented from producing alternative interpretations of EU law, so that EU law is interpreted uniformly across the block;
    • the existence of such a doctrine also means that the legislative bodies of the individual Member States are prevented from enacting legislation that would conflict with EU law, so that EU law is applied in a consistent and coherent manner in the domestic systems of the Member States.

However, as has already been noted, there is nothing in the Treaties that specifically states that EU law, in whatever form, takes precedence over national law. The closest the Treaties come is in the principle of sincere co-operation contained in Article 4(3) TEU (formerly so-called duty of loyalty in Article 10 EC Treaty) which can be paraphrased in the following terms:

The Court of Justice, then, has been proactive and instrumental in ensuring that the objectives of the Treaties are achieved in the Member States, and in doing so supremacy has been a powerful tool in its hands.

Quite simply, the full economic integration necessary for the achievement of a Single Market would have proved unmanageable if it had been possible for Member States to ignore or even deliberately defy the supranational powers of the institutions of the EU.

Taking a long-term view, probably the most logical basis for supremacy is that without it a full integration, including the political integration required for a federalist system, would also prove impossible.

Whilst the abandoned Constitutional Treaty included an EU law primacy clause, this provision was not included in the Treaty of Lisbon as it proved too controversial for many Member States. Following the Inter-Governmental Conference 2007 on the adoption of the Lisbon Treaty, the European Council acknowledged the doctrine of primacy in Declaration 17 appended to the Treaty of Lisbon and which states that:

in accordance with well settled case law of the Court of Justice of the European Union, the Treaties and the law adopted by the Union on the basis of the Treaties have primacy over the law of Member States, under the conditions laid down by the said case law.

Compared to protocols and annexes which have legal force, declarations attached to the Treaties are merely politically binding. Still, this political declaration validates the ECJ jurisprudence and reinforces the fundamental principle that where EU law applies, EU law will prevail over national law.

7.2 The development of a doctrine of supremacy

7.2.1 The early definitions of ‘supremacy’

The earliest indication of a concept of supremacy/primacy comes from the judgment of the ECJ in the case of Van Gend en Loos v Nederlandse Administratie der Belastingen (Case 26/62) [1963] ECR 1; [1963] CMLR 105. The case involved an introduction of a Dutch law which was contrary to obligations imposed on Member States in an EC Treaty Article. Because the Article implicitly conferred rights on individuals and the litigant suffered financial loss as a result of the Dutch law, an Article 177 (now Article 267 TFEU) reference posed the question for the ECJ of whether or not the Treaty created rights on behalf of individuals that national courts must then protect.

In the reasoned opinion the Advocate-General declared that this was not the case and that the appropriate means of resolution in such circumstances was in an action against the Member State under Article 169 (now Article 258 TFEU). The judges in the ECJ disagreed and in their judgment identified:

This is not a complete definition of ‘supremacy’. Nevertheless, the ECJ avoided the problems that would go with classing EU law as international law, to which different Member States, because of their different constitutions, would adopt completely different approaches. What the ECJ identified in the case was that the EU must be viewed as a different and unique legal order distinct from either national law or international law because of the transfer of powers from the Member States over to the EU institutions. Therefore, the key point was the clear statement that, by accepting entry into the (then) Community, the Member States were limiting their otherwise sovereign rights to legislate contrary to the requirements of EU law in those areas affected by it.

The ECJ had to wait only two years to produce a more complete definition and more extensive explanation. This came in another reference (now under Article 267 TFEU), this time from an Italian court. Again, it involved legislation that was passed after accession to the Treaties that was inconsistent with provisions of EU law.

The Court also continued by identifying the clear consequence of the transfer of powers:

In reaching its decision the Court identified that supremacy was confirmed by the wording of Article 288 which refers to the concept of direct applicability (see section 3.2) and also the binding nature of Community legislation. In its conclusion it explains fully its reasoning:

Three clear propositions emerge from the statements in these two cases:

  • the Member States, by joining the EU, had willingly transferred some of their national competence and thus given up certain of their sovereign powers to make law on certain issues;
  • both the Member States themselves as well as their citizens are bound by EU law;
  • the Member States, as a result, cannot unilaterally introduce new national laws that would then contradict EU law.

Supremacy or primacy?

Whilst the majority of EU law textbooks will use the term ‘supremacy’, it is not the one chosen by the ECJ, the drafters of the former Constitutional Treaty or the Treaty of Lisbon. It is worth noting that the EU has 24 official languages across the 27 Member States. It appears that any use of the word ‘supremacy’ has tended to be found in the English translation of two ECJ case laws (Wilhelm (Case 14/68) [1969] ECR 1 and Fratelli Variola (Case 34/73) [1973] ECR 981) – whereas the ‘primacy’ terminology seemed to be favoured in the language versions of these judgments.

The term ‘supremacy’ evokes a state of superiority or hierarchical relationship whereby EU law would have a ‘higher’ ranking or status over national law. ‘Primacy’, by contrast, denotes a more pragmatic and jurisdictional relationship – it is about determining which law (EU or national) shall have precedence in a given situation by establishing which jurisdiction/competence (either EU or national) governs that situation.

Avbelj (2011) suggests that the difference between supremacy and primacy matters as, more than a question of semantics, it reveals how the doctrine shapes the EU legal order. The distinction between conceptualisations is pertinent to the nature of the EU and to the relationship between EU law and its application in the domestic legal systems of the Member States.

7.2.2 The wider application of the doctrine of supremacy

The ECJ has had the opportunity to restate the principle on a number of occasions. On one occasion it was able to expand by explaining that EU law cannot be invalidated by any measure of national law, from whatever source.

The message from the ECJ was very clear: in determining any ‘pecking order’ of laws, EU law takes precedence over even the constitutions of the Member States.

In further restating the principle of supremacy the ECJ has subsequently made it abundantly clear that, in the event of any conflict or inconsistency between national law and EU law, the domestic court has an absolute requirement to give effect to EU law over any conflicting law, whatever the date of passing of that law.

The ECJ in its response made the following observations:

In its conclusions the ECJ again clearly stated the consequences for Member State courts:

Over time, EU law has been held to have primacy over conflicting national law in many cases. What the cases clearly identify is that supremacy applies whatever the particular type of EU law. We have already seen the applicability of the doctrine in relation to Treaty Articles and also Regulations. EU law has also been held to have primacy in the case of:

  • Directives – Becker v Finanzamt Munster-Innenstadt (Case 8/81) [1982] ECR 53, [1982] 1 CMLR 499 (where the applicant was then permitted to enforce a VAT Directive against the German tax authorities);
  • Decisions – Salumijico di Cornuda (Case 130/78) [1979] ECR 867;
  • general principles of law – Wachauf v Germany (Case 5/88) [1989] ECR 2609;
  • international agreements made between EU and non-EU states – Nederlandse Spoor-wegen (Case 38/75) [1975] ECR 1439.

The doctrine of supremacy also will apply not merely where there is national law that directly conflicts with a provision of EU law. The doctrine will apply also in circumstances where a contradictory provision in national law, while not directly conflicting, nevertheless by its existence encroaches on the field of EU legislative powers.

7.2.3 The extreme consequences of a doctrine of supremacy

The ECJ has had many opportunities to state and restate the rule. This occurred not just in the early years of the Treaties but has surfaced in recent times too.

Possibly the most far-reaching application of the rule came in a case involving the UK when it was a Member State of the EU. The case was dramatic because in the most positive terms it conflicted with the notion of parliamentary supremacy within the English Constitution, and directed English judges to exercise a power that would have formerly been considered to be beyond them: to suspend operation of and in effect declare invalid provisions of an Act of Parliament.

In its conclusions the ECJ was clear on the course of action to be taken by the national court and the effect of supremacy on the situation:

Factortame (1990) then represents a major statement of supremacy by the ECJ over national courts and of EU law over national law and demonstrates the supranational power of the institutions of the EU.

Overview of key principles

Case law Key principles
Van Gend en Loos Community/Union is a new legal order
Costa v ENEL Supremacy of EC/EU law on three grounds: independence, uniformity and effectiveness/efficacy
Handelsgesellschaft Unconditional and absolute nature of the principle Applies to Treaty and all the acquis communautaire
Simmenthal EC/EU law prevails over previously and subsequently enacted national laws Conflicting domestic law must beset aside and EC/EU provision applied
Factortame Conflicting domestic legislation can be temporarily set aside Interim can be granted until preliminary reference rendered

7.3 Supremacy/Primacy and Member States

The above historical development of EU law primacy specifically refers to the jurisprudential creation of the doctrine through the decisions of the ECJ. In other words, the justifications and principles established for the primacy of EU law over domestic law follows the rationale and narrative as set by the Court of the Justice of the European Union. From that perspective, the doctrine is therefore absolute for the efficacy and uniformity of the EU law to be maintained.

However, there is another perspective, and therefore a different narrative on the meaning and scope of the doctrine – and that narrative is one of ‘relative’ primacy, as expressed by the various national courts of the Member States.

Several domestic constitutional courts have indeed raised and explored constitutional questions arising from the interpretation, application and development of the doctrine of EU law primacy over national law. Consequently, domestic constitutional courts have developed their own rationale for adopting the doctrine of primacy of EU law, given that the nature and extent of the constitutional issues arising would depend on the character of the particular constitution in question, and whether the national legal system would be a monist of dualist system.

A flow diagram depicts the development of a definition of the supremacy of European Union law. The flow starts with the Treaty, which creates a new legal order, leads by carrying a clear limitation, and continues with the validity of a community measure. It finally leads to the requirement that every national court must, in a case within its jurisdiction, apply community law in its entirety.

Figure 7.1 The development of a definition of supremacy of EU law

In the BENELUX countries, for instance, the questions have been limited and those Member States have been able to accept EU law primacy, with little difficulty, because of the general effect of international Treaties on their constitutions. By contract, in Member States such as France, Germany and Italy, there have been constitutional issues leading to cause for concern for national courts.

7.3.1 Belgium and supremacy of EU law

The Belgian Constitution operates according to a ‘monist’ system (in contrast to the dualist constitution of the UK, for instance). In simple terms, a monist constitution operates so that international treaties automatically become part of the national legal order on signing of the Treaty. In contrast, a dualist approach requires that, besides becoming a signatory to a Treaty, the state in question is required to ratify and incorporate the provisions into its national law by further legislation. This was, for instance, the reason the UK passed the European Communities Act 1972 when it first joined the Union.

In Belgium, then, the Treaties automatically became incorporated in national law once entry into the EC was achieved by signing the Treaty of Accession.

The issue of supremacy was in any case settled in subsequent case law:

7.3.2 France and supremacy of EU law

In France, whilst EU law primacy is now an accepted doctrine, it has not been without debate between the three highest courts of the French legal order: Conseil Constitutionnel (constitutional court); Cour de Cassation (supreme ordinary court); and Conseil d’État (supreme administrative court). There was a division that initially manifested between the first two courts, who accepted the doctrine, and the Conseil d’État who initially refused to embrace the primacy of EU law.

At quite an early stage, the Cour de Cassation in Von Kempis v Geldof [1976] 2 CMLR 462 accepted both the reasoning and the case law of the ECJ in declaring that EU law would take precedence over French legislation. This was a reaffirmation of the Jacques Vabre case [1975] 2 CMLR 336, which was subsequently endorsed by the Conseil Constitutionnel in 1986.

By contrast, the Conseil d’État, in Minister for the Interior v Cohn-Bendit [1980] 1 CMLR 543, on the matter of Directives issued under Article 39 (now Article 45 TFEU), stated that a Directive could not be used as a means of challenging internal administrative law. This created a split between the courts and uncertainty within the French legal system, which was not fully accepting of EU law primacy over conflicting French domestic law.

In Nicolo [1990] 1 CMLR 173, the Conseil d’État cautiously accepted EU law primacy by first assessing the compatibility of EU law with French law; and then by relying Article 55 of the French Constitution in asserting that French courts are enabled to set aside domestic law contrary to the Treaties. In other words, French courts accept the primacy of EU law only because EU law first conforms with the provisions of the French Constitution.

The Conseil d’État eventually expressly reversed its Cohn-Bendit jurisprudence and fully embraced EU law primacy in Mme Perreux (2009). In this case, a French national, who applied for a post at the École de Magistrature, claimed discrimination on account of her activities for the Union of Judicial Authorities. She invoked an EU Directive that had not been implemented by France, against the domestic administrative act she challenged. 30 years on and after an evolutive case-law, the Conseil d’État decidedly resolved the split by recognising explicitly both the primacy of EU law and the direct effect of Union law.

7.3.3 Italy and supremacy of EU law

The Italian constitutional court initially took quite a hard line, as was seen in Costa v ENEL (1964). At that point, the view clearly expressed was that Italian legislation coming after the Treaty must be followed, and this was the argument presented in the reference at that time.

However, the court then moved sharply away from that position, accepting the supremacy of EC (now EU) law:

Since that time Italy has been firm in its observance of the principle of supremacy. It has tended to operate towards a constructionist approach: that construction of Italian law must be consistent with the demands of EU law.

7.3.4 Germany and supremacy of EU law

Initially, Germany was unwilling to accept supremacy as an absolute principle because of the potential conflict with the protections of human rights contained in the German Constitution. This position was clearly demonstrated in the position taken by the German court in International Handelsgesellschaft GmbH v EVGF (1970) discussed in section 7.2.2. It was also discussed at length in Wunsche Handelsgesellschaft [1987] 3 CMLR 225. Here, the German constitutional court reversed its position in the former case and accepted the principle of supremacy provided that the EU law could guarantee the protection of human rights. In a different context, in Kloppenberg [1988] 3 CMLR 1, where the tax courts had originally held against the direct effect of a Directive on VAT, the constitutional court reversed this and upheld the supremacy of EU law in clear terms.

There is, then, some discrepancy between the attitude of different courts to supremacy. This debate has continued. In Brunner v The European Treaty [1994] 1 CMLR 57 the applicant had challenged the constitutional legitimacy of Germany signing the Treaty on European Union. The German constitutional court disagreed and upheld the validity of German membership of the new Treaty. Nevertheless, the judges passed significant comment on the relationship between the German state and the EU:

In the same vein as the above, in the Treaty of Lisbon judgment [2 BvR 182/09] pronounced 30 June 2009 the German constitutional court confirmed the conditionality of the doctrine of EU law. The Court stated that it has the competence to review the compatibility of EU law, including EU Treaties amendments and transfer of powers from Germany to the EU, should that threaten the identity of the German constitution.

The ECJ’s stance that EU law has primacy and the doctrine is absolute is based on the premise that the EU is sui generis – a unique supranational organisation with its own institutions and competences following a transfer of powers from the Member States (thus limiting their individual/unilateral action). However, it is exactly that last point which underscores the argument of the German constitutional court. The German Court argues that, far from being absolute, EU law primacy is conditional because the powers of the EU and its institutions derive from a (partial) transfer of national sovereignty. In other words, the authority of the EU was given by the Member States, ergo the primacy of EU law is only valid insofar as it is permitted by their national constitutions. This constitutional conditionality theory of EU law primacy was also confirmed in the Danish Maastricht Treaty Ratification judgment [1999] 3 CMLR 854, and is often referred to as ‘kompetenz-kompetenz’. Following that school of thought, if an act of the EU is deemed to operate outside of its remit, or the Court of Justice acted ultra vires, it may be challenged by the domestic court.

This is exactly what happened in 2020 when the German Constitutional Court ruled that a measure of the European Central Bank (a bond-buying programme) was potentially illegal; and that the Court of Justice of the EU had acted ultra vires in approving the scheme.

Following the GCC decision, the European Commission launched legal proceedings against Germany in June 2021, by sending a letter of formal notice under Article 258 TFEU for infringement of the primacy of EU law. The proceedings were eventually closed, once Germany issued assurances that EU law primacy is and would be respected.

7.3.5 Poland and EU law primacy

The Polish Constitutional Tribunal has accepted the principle that EU law may have primacy over conflicting domestic law on a conditional basis, i.e. with inherent constitutional limitations applying to the principle. In several judgments, the PCT has reiterated the primacy of the Polish constitution as the supreme law of the Republic of Poland, specifically in relation to international agreements, transfer of powers and EU law (K 18/04 (2005); P 37/05 (2006); K 32/09 (2010)).

The Polish acceptance of the doctrine therefore echoes the French and German constitutional courts’ rationale that it is conditional on the compatibility of EU law with the provisions of their national constitution. In particular, Article 8(1) of the Polish Constitution affirms that it is supreme law over which EU law may not have precedence.

More controversially, however, in a judgment delivered on 7 October 2021 (K3/21), the Constitutional Tribunal assessed the conformity to the Polish Constitution of selected provisions of the Treaty on the European Union, and concluded that these TEU provisions were incompatible with Polish Constitutional provisions.

In the context of what is a termed a ‘political democratic back-sliding’ on the application of the rule of law in Poland, this judicial development raises serious questions about the future of European integration and of Poland within the European Union.

Further reading

Articles

  • Arena, A, ‘From an Unpaid Electricity Bill to the Primacy of EU Law: Gian Galeazzo Stendardi and the Making of Coasta v ENEL’ (2019) 30(3) Eur J Int Law 1017–1037.
  • Avbelj, M, ‘Supremacy or Primacy of EU Law – (Why) Does it Matter?’ (2011) 17 ELJ 744.
  • Charpy, C, ‘The Conseil d’État Abandons Its Cohn Bendit Case-Law; Conseil d’État, 30 October 2009, Mme Perreux’ (2010) 6 Eur Const Law Rev 123–136.
  • Gábor, H, ‘Conclusive Remarks, Constitutional Adjudication in Europe between Unity and Pluralism’ (2018) 2(2) Ital J Public Law 477–484.
  • Tuominen, T, ‘Reconceptualizing the Primacy-Supremacy Debate in EU Law’ (2020) 47(3) Leg Issues Econ Integration 245–266.

Books

  • Alter, KJ, Establishing the Supremacy of European Law: The Making of an International Rule of Law in Europe (OUP, 2003).
  • Craig, P and de Burca, G, EU Law: Text, Cases and Materials (OUP, 2020).
  • Kaczarowska-Ireland, A, European Union Law (Routledge, 2016), chap. 9.
  • Schütze, R, European Union Law (OUP, 2021).

8 The relationship between EU law and national law – direct effect

DOI: 10.4324/9781003218562-8

8.1 Introduction

Where primacy of EU law articulates the relationship between Member States and the EU within a supranational context, direct effect triangulates this relationship by establishing a relationship between EU citizens and the Union anchored in its legal order. Direct effect – in focusing on granting individual rights under EU law – is an additional constitutional pillar that ensures EU law is applied harmoniously throughout the Member States.

Primacy gives practical effect to the concept of supranationalism by ensuring that the Member States cannot put national self-interest before the law of the EU. Direct effect demonstrates the supranational nature of EU law by ensuring that citizens can invoke EU law and therefore have it enforced before the domestic courts of the Union Member States.

Direct effect was not created or anticipated by the Treaties; and it is thus not defined within them. The doctrine of Direct Effect (and its related principles of Consistent Interpretation and State Liability) is entirely a creation of the Court of Justice of the European Union (CJEU/ECJ). Whilst it does in some ways relate to other concepts identified in the Treaties, it must be distinguished from them. Article 288 TFEU, in establishing and defining the various types of legislation, for instance refers to the following.

8.1.1 General applicability

A measure of general applicability applies universally, i.e. in a generic manner, across the Member States of the EU. For instance, a Regulation is described as being generally applicable because it applies to unidentified, abstract categories of natural and legal persons; whereas a Decision could never be generally applicable, because it is addressed to a specific party (e.g. to a Member State in a particular).

8.1.2 Direct applicability

Direct applicability refers to the automatic legal standing of an EU measure within the domestic legal system of a Member State. A directly applicable EU measure, adopted by the Council and the European Parliament through either the ordinary or special legislative procedure, automatically becomes law in the Member States once passed. For instance, a Regulation is described as directly applicable in Article 288 TFEU, and as such it automatically becomes law in the Member States, i.e. there is no requirement for implementing legislation by the Member States. What may be appropriate is for them to produce law which repeals what would otherwise be inconsistent national legislation. Nevertheless, applicability in this context means that the Regulation would apply despite the conflicting national measure, on the basis of the doctrine of EU law primacy, whereby the latter must prevail over conflicting national legislation where EU law applies. A Directive, by contrast, is only binding as to the goals to be achieved, which means that Member States have discretion on how to achieve these goals, Therefore, a Directive requires implementation by Member States within a set period (usually two years) and cannot, therefore, be directly applicable.

Direct effect, by comparison, refers to the actual enforceability of EU law, of whatever type, in the national courts of Member States. Where EU law creates obligations, it therefore creates rights – this concept of obligations/rights enforceability can be applied to Treaty Articles, Regulations, and even to Decisions. However, the concept of direct enforceability or effect is much more problematic when it is applied to Directives.

Inevitably, this means that there are significant distinctions between direct applicability and direct effect:

  • A measure may be directly applicable without necessarily creating rights that are enforceable. Indeed, whilst an EU measure can apply in a legal system, like a procedural Regulation for instance, it does not necessarily create individual rights. Therefore, an EU measure can be directly applicable yet not have direct effect.
  • A measure can create individual rights and therefore be directly effective, although it may not be directly applicable. This is the case with Directives, but only in limited circumstances.
  • A measure could also be both directly applicable and directly effective.

Therefore, whilst direct applicability relates to the automatic legal force of a given EU legislation (i.e. it is part of the domestic legal system and body of law), direct effect is concerned with the enforceability of a specific provision/obligation/rights within that measure (i.e. whether individuals can invoke that legislation in domestic judicial proceedings to ensure the enjoyment of the rights it bestows upon them).

There is no doubt that direct effect is one of the most important legal developments of the ECJ. It has been instrumental in ensuring that the broad objectives of the Treaties have been observed by the Member States and incorporated into national law.

Various academics have commented on the significance of the concept: ‘[Direct effect is] the first step in the judicial contribution to federalism’ (P Craig, ‘Once Upon a Time in the West: Direct Effect and the Federalisation of EC Law’ 1992 OJLS 453), ‘a second principle of western jurisprudence to run alongside supremacy; namely the rule of law’ (I Ward, A Critical Introduction to European Law [Butterworths, 1996], p. 57), and ‘[The most powerful justification for the creation of the concept is that it] enhances the effectiveness or effet utile of binding norms of Community law’ (J Shaw, European Community Law [Macmillan Professional Masters, 1993], p. 151).

As with primacy of EU law, one of the most interesting aspects of the doctrine is that it is entirely the creation of the ECJ. It is difficult to imagine where the EU would have stood today without the ECJ taking such an uncompromising stance in creating through primacy/supremacy and direct effect the means by which the broad objectives of the Treaties can become established in national law. Inevitably, the uncompromising nature of the Court has at times resulted in conflict with Member States. This in turn has led to even more uncompromising behaviour by the ECJ in developing the principles of indirect effect and state liability.

Nevertheless, the CJEU/ECJ is dependent upon the national courts to develop the doctrine of direct effect and any other principle of EU law. Indeed, the nature of the Union is to operate on a collective international agreement (Treaties) through which Member States have transferred selected powers to EU institutions; and Article 267 TFEU (preliminary ruling procedure) in particular, is the means through which national courts refer questions of EU law interpretation, application, and validity to the CJEU. As such, most cases creating EU law principles, such as primacy and direct effect, are brought forward by the national courts, who are then the enforcers of EU law. F Mancini identifies the significance of this:

The national courts … by referring to Luxembourg sensitive questions of interpretation of Community law … have been indirectly responsible for the boldest judgments the Court has made. Moreover, by adhering to these judgments in deciding the cases before them, and therefore by lending them the credibility which national judges usually enjoy in their own countries, they have rendered the case-law of the Court both effective and respected throughout the Community.

F Mancini, ‘The Making of a Constitution for Europe’ (1989) 26 CMLR 595

8.2 The concept of direct effect

8.2.1 The origins of direct effect

To reiterate, the doctrine of direct effect is not mentioned anywhere in the Treaties; instead it has been extrapolated from the Treaties and developed through judicial decisions of the CJEU. In fact, the CJEU had the opportunity to consider the enforceability of EU law first in the same case in which it also effectively developed the doctrine of supremacy/primacy.

Inevitably, the two concepts – primacy and direct effect of EU law – go hand in hand, often referred to as the constitutional pillars of EU law. They are essential for the ultimate success of the Treaties and the principle of supranationalism upon which the Common Market is based.

Direct effect was relevant to the case below because there was a conflict between what fell within national law and how it conflicted with EU law. The citizen would have had no protection if he had been unable to rely on EU law in the case.

The ECJ pointed out then that, since the Treaty was clearly intended to affect individuals as well as Member States, it must also be capable of creating rights that would also be enforceable by individuals within the national courts.

In this way the Court concluded that since Article 12 (now Article 30 TFEU):

The Court was also conscious of the limited value of the course of action that had been proposed by the Advocate-General, an action against the Dutch State under Article 258. The Court explored this possibility and gave its reasons against it.

The original method used to ensure that EU law was in fact enforced in Member States was by an action against the state through the enforcement proceedings in Articles 258 and 259 TFEU. As seen in the passage above, the ECJ established the concept of direct effect because it was a more effective means of ensuring that citizens of Member States could enforce the rights given to them by the Treaties than the traditional means.

8.2.2 The criteria for direct effect

Of course, the effect of Van Gend en Loos (1963) was initially quite limited in that direct effect only had to be applied to what are often referred to as ‘standstill’ Articles, or in other words prohibitive Treaty Articles. In addition, the action was in any case against the state itself.

The ECJ was able to develop the doctrine in Van Gend (1963) because the Article in question involved an obligation on the part of the Member States not to increase existing Customs duties or indeed create new ones. In this way the ECJ could quite easily justify enforcement of the provision because it was, as the Court described it, ‘clear, precise and unconditional’ and did not depend on any further action being taken for its implementation (self-executing), either by the Community institutions or by the Member States themselves – this is the original formulation of the Van Gend en Loos criteria.

However, as more cases arose, which besides Treaty provisions would involve Regulations, Decisions and Directives, the inherent limitations to the Ven Gend en Loos formula would progressively disappear in later CJEU judgments. For instance, the self-executing criterion would later be dropped in cases involving Directives (because these are not self-executing, they must be implemented by national law). By accommodating the application of the doctrine to wider EU measures, the CJEU has ensured the extension of the scope of direct effect. In this way, not only could a wider range of EU law be enforced, but EU measures could also be directly effective against other individuals, as well as invoked against the state.

Later judgments did confirm the criteria for establishing direct effect developed by the Court in Van Gend (1963), for instance in Reyners v Belgian State (Case 2/74) [1974] ECR 631. The criteria for direct effect are nevertheless most commonly referred to as the Van Gend en Loos criteria or formula:

  • The provision must be sufficiently clear and precisely stated – the focus is on the wording here. A sufficiently clear and precise provision contains a clearly and precisely stated prohibition, obligation and/or right. For instance, the ECJ in Defrenne v SABENA (Case 43/75) [1976] ECR 455 was satisfied that the principle that ‘men and women shall receive equal pay for equal work’ was sufficiently precise (i.e. right to equal pay) to create direct effect, even though the exact meaning of ‘equal pay’ and ‘equal work’ would inevitably require further definition by the courts.
  • The provision must be unconditional or ‘non-dependent’ – there are no ‘ifs’ here, in the sense that the obligation or right contained in the EU provision should not be subject to any institutional, structural, or other condition for its application; e.g. it should not depend on the intervention of another body or require further legislative action either by the Community institutions or by Member States.

Added to this, of course, for individuals to gain enforceable rights from the provision:

  • There must in fact be an identifiable right granted by the Treaty or legislative provision and on which the citizen can then rely (Francovich v Italy (Cases C–6 and 9/90) [1991] ECR I–5357). Following this reasoning, when the party liable for the prohibition and/or obligation and the beneficiary of the related right are clearly and precisely stated in the provision, then a right can be identified.

Not all provisions do conform to the criteria, even though they may appear to be worded in unconditional terms. In Casati (Case 203/80) [1981] ECR 2595 the ECJ was concerned with whether the provision under the old Article 71 (now repealed) that ‘member states shall endeavour to avoid introducing … new exchange restrictions’ was directly effective. The Court held that the words ‘shall endeavour’ were in fact insufficient to create unconditional obligations and that therefore the provision could not be directly effective. By contrast, words that would convey unconditionality would for instance include a modal verb such as ‘must’. Generally, it can be argued that determining whether a provision is direct effectively is in large part a semantic exercise.

8.2.3 Vertical direct effect and horizontal direct effect

Because the action was against the state, the case of Van Gend en Loos (1963) addressed a vertical relationship, i.e. an individual against the state, and confirmed it is possible for an individual to invoke a provision of EU law against the state. Therefore, the case did not deal with the issue of whether or not a citizen could rely on the principle of direct effect to enforce a provision against another citizen.

This was one of the questions for the Court in Defrenne v SABENA (1976), which involved a claim for equal pay made against an employer under Article 157 (see Chapter 18). The ECJ rejected the argument that direct effect was a means only of enforcing substantive EU laws against the Member States. The Court identified that there were two types of direct effect: vertical direct effect (between an individual and the state) and horizontal direct effect (between two private parties, i.e. individuals). This was necessary because otherwise citizens would be denied effective remedies where they were granted rights under EU law.

The Court, in essence, then, also clarified the distinctions between the two:

Vertical direct effect in one sense concerns the relationship between EU law and national law:

  • measures of EU law create obligations on the state to ensure their observance;
  • a failure on the part of Member States to honour Treaty obligations would usually give rise to an action against the state under Article 258 TFEU;
  • but vertical direct effect means that the process is not necessary since an individual can rely on the measures in an action against the state (as was the case in the original Van Gend en Loos case and in both Van Duyn v Home Office (Case 41/74) [1974] ECR 1337 and Pubblico Ministero v Ratti (Case 148/78) [1979] ECR 1629 which were both fought against government departments);
  • indeed the principle has been extended to cover other ‘public bodies’ other than the state itself, and these are known as ‘emanations of the state’.

Horizontal direct effect, on the other hand, is concerned instead with the relationship between individuals and other individuals (private parties):

  • this could include any private body including companies;
  • where a measure is horizontally directly effective, it creates rights between citizens and is therefore enforceable by them in national courts;

However, horizontal direct effect mainly applies to Treaty provisions and Regulations. Inevitably, given the nature of a Directive, the ECJ has stated that Directives do not have horizontal direct effect; the main reason being two-fold: Directives are addressed to and impose obligation upon Member States (not individuals), and they are subject to domestic implementing measures.

Given this restriction, the application of the doctrine of direct effect with regards to Directives has proven complex; therefore the ECJ has been active in ensuring that citizens will not lose out merely because the provision in question is a Directive, by creating the principles of consistent interpretation and incidental effect for instance.

The relationship between horizontal direct effect and vertical direct effect can be shown in diagram form as in Figure 8.1.

A flow diagram depicts the relationship between vertical direct effect and horizontal direct effect. The flow starts from European Union Law, which can create rights on which nationals of member states can rely, which leads to European Union citizens, which leads to individuals. Creates obligations on member states through Treaty Articles, and legislation leads to member states. The horizontal and vertical direct effects are represented with horizontal and vertical dotted boxes.

Figure 8.1 The relationship between vertical direct effect and horizontal direct effect

8.3 The application of direct effect

The Court of Justice has developed the principle of direct effect so that it applies generally to most types of EU law, both primary and secondary. It has generally followed its own criteria in Van Gend en Loos (1963). However, although the test from the case was originally strictly applied, the CJEU has also gradually taken a more relaxed approach to ensure that citizens can take advantage of the rights given to them in the Treaties.

There are a number of consequences of this relaxation:

  • the Court has in effect assumed responsibility for ensuring the effective integration of EU law;
  • provided that the criteria are generally met, then direct effect of a substantive measure is almost assumed;
  • the direct effect of a provision is only likely to be denied when there would be serious political or social consequences;
  • as a result, it can be argued that direct effect has become essentially a question of policy for the ECJ (as illustrated by the developments concerning the EU Charter of Fundamental Rights for instance).

8.3.1 Direct effect and Treaty Articles

The concept of direct effect was first accepted in Van Gend en Loos (1963) and this case of course involved a Treaty Article, Article 12 (now Article 30).

This was a so-called standstill Article. The principle of direct effect was also applied in respect of Treaty Articles imposing a duty on Member States to act, which would have been the case with removal of features from their national laws during the transitional period that would be detrimental to the attainment of the Common Market.

The principle, however, has been extended to cover all of the main substantive Treaty Articles. So, for example, it has been applied to both Article 34 and Article 35 on the free movement of goods in Ianelli & Volpi SpA v Meroni (Case 74/76) [1977] ECR 595 (see Chapter 14).

It has also been applied to the free movement of workers under Article 45 TFEU in Van Duyn v Home Office (1974); to rights of establishment under Article 49 TFEU in Reyners v Belgium (1974); and to the freedom to provide services under Article 56 TFEU in Van Binsbergen v Bestuur van de Bedriifsvereniging voor de Metaalnijverheid (Case 33/74) [1974] ECR 1299 (see Chapter 13). Both Article 49 and Article 56 TFEU envisage the introduction of Directives for the harmonisation of qualifications throughout the EU. The argument that this made the Articles conditional and dependent and that they could not as a result be directly effective was expressly rejected in the Reyners (1974) case.

Besides this, the principle has also ensured enforceability of EU competition law under Article 101 and Article 102 TFEU in Brasserie de Haecht SA v Wilkin-Janssen (No 2) (Case 48/72) [1973] ECR 77 (see Chapter 16).

We have also seen already that the principle of equal pay for equal work in Article 157 was declared directly effective in Defrenne v SABENA (1976). The Court did, however, take the unique step of declaring the Article only prospectively directly effective.

The role of the ECJ in declaring the major EU policies directly effective has clearly been critical in defining those Treaty provisions as well as in ensuring that citizens can rely on them. One further point to add, which is of course apparent from the cases above, is that a Treaty Article will be both vertically and horizontally directly effective.

Charter of Fundamental Rights

The CFR is primary legislation that has legal force and the same constitutional value as the Treaties. Therefore, the applicability principles of the doctrine of direct effect for the CFR are the same as for Treaty provisions – albeit within the scope of application of Article 51 CFR, which establishes that the Charter applies to institutions and bodies of the European Union, and to Member States in the implementation of EU law.

CFR provision can thus have vertical direct effect, as individuals can invoke their CFR rights against a Member State (C–617/10 Åklagaren v Hans Åkerberg Fransson) or an institution of the Union (Cases C–293 and 594/12 Digital Rights Ireland v Minister for Communications et al), subject to article 51 CFR limitations and Van Gend en Loos criteria.

By contrast, the horizontal effect of the CFR, i.e. the ability for an individual to invoke the CFR against another individual, is more contentious given that the Charter is addressed to EU institutions and Member States, upon whom obligations to respect and protect fundamental rights are imposed. Still, since it gained legal binding force in 2009, the CJEU/ECJ has gradually found that the CFR can produce horizontal effect in some proceedings between private parties.

Controversially in C–144/04 Mangold (2005), the CJEU recognised that the principle of non-discrimination included in the general principles of EU law (GPEUL; i.e. the fundamental rights mechanism under EU before the Charter of Fundamental Rights), which was included in Directive 2000/78, could be invoked in horizontal relationships (thus allowing that part of the directive to have horizontal direct effect). This decision did raise questions as it seems to imply that directive could have direct horizontal effect – even before the implementation period of the Directive has lapsed.

This line of argument was later refined post-2009, after the Charter of Fundamental Rights was accorded legal binding force. In C–555/07 Kücükdeveci v Swedex, the non-discrimination provision contained in the CFR (also GPEUL), was given direct horizontal effect and primacy over a conflicting domestic provision, where its application falls within the scope of EU law. The CJEU clarified that whilst the principle of non-discrimination was given horizontal direct effect, that is not the case for the Directive itself – a nuanced yet significant distinction.

Whilst the CFR can have direct effect (both vertical and horizontal), the CJEU reiterated in C–176/12 AMS that for such a provision to have direct effect it must still fulfil the Van Gend en Loos criteria. In this case, the wording of Article 27 CFR was not sufficiently clear and precise to create an obligation. Nevertheless, once the Van Gend en Loos criteria are met, the CJEU has determined that certain CFR provisions (e.g. Articles 21, 31(2), 47) can have horizontal direct effect – this was revealed through a series of judicial decisions in 2018 including: C–414/16 Egenberger; C–68/17 IR v JQ; C–193/17 Cresco Investigation; C–569/16–570/16 Bauer & Willmeroth; C–684/16 Max-Planck.

8.3.2 Direct effect and Regulations

The definition of ‘Regulations’ under Article 288 TFEU means that they are both of ‘general application’ and ‘directly applicable’. The consequence of this is that they may also create obligations without need for further enactment by either the institutions or the Member States.

It therefore follows that they are also capable of direct effect provided that they satisfy the Van Gend en Loos (1963) criteria. On the basis that a Regulation is directly applicable, it can never be construed as conditional; therefore, provided it contains a recognisable right, the only real test for the ECJ is whether or not the provision is stated in sufficiently clear and precise terms. However, it naturally follows that where a Regulation is too vague in its terms, then it may not be directly effective.

Again, where the ECJ accepts the direct effect of a Regulation, then direct effect is both vertical and horizontal, as was confirmed in Muñoz (Case C–253/00) [2002] ECR 7289.

8.3.3 Direct effect and Decisions

As we have seen, a ‘Decision’ is defined in Article 288 TFEU as being ‘binding in its entirety on the party to whom it is addressed’.

Since Decisions generally are addressed to parties who are in breach of EU obligations, there is a clear need for the citizens who suffer as a result of those breaches to be able to rely on the decisions. The ECJ has itself accepted that.

The ECJ concluded that it could and identified that it would be:

Of course, once again, all of the criteria for direct effect would need to be present. The above case law, which relates to an instance where a decision was combined with a Treaty provision and directive, was later confirmed in a case which only involved a decision (Hansa Fleisch (Case C–156/91) [1992] ECR I–5567).

The fact that decisions are binding only on the party to whom the decision is addressed means that there is not the same problem of lack of direct applicability that occurs with Directives. Nevertheless, some of the problems that occur with Directives in relation to private parties could occur where that party is not the one to whom the decision is addressed (e.g. where a decision is addressed to a Member State for instance).

8.4 The issue with the enforceability of Directives

8.4.1 Direct effect and Directives

Again, the character of a Directive is defined in Article 288 TFEU. Under the Article, Directives are said to be ‘binding as to the result to be achieved’ but ‘leave to the national authorities the choice of form and method’. This means that Directives have specific objectives which create legally binding obligations on Member States to pass national laws within a set time (usually within a two-year window) to achieve those objectives.

In this way Directives cannot in themselves automatically create substantive rights that citizens are then able to enforce.

  • First, they are not directly applicable, i.e. they are not self-executing. They do not automatically become laws in the Member States. On the contrary, the Member States are allowed to achieve the objects contained in the Directives in whatever way they choose.
  • Second, because they arguably fail one part of the test in Van Gend en Loos (1963), i.e. they are entirely dependent on implementation by the Member States, therefore until the two-year transposition period lapses, they are conditional.

The problem first came to the attention of the ECJ the Van Duyn (1974) case, a case that raised the question of the direct effectiveness of a Directive in a vertical relationship. Although this was a somewhat reversed situation, as in this case it was the Member State who invoked the Directive and pleaded it was enforceable. The important point for the citizen involved was whether the procedural safeguards contained in the Directive bound the Member State.

The ECJ was thus prepared to overlook the potential limitation involved in the definition of ‘Directives’ given by Article 288 TFEU, again to ensure the effet utile (usefulness) of the measure. The Court foresaw that this type of legislation could be made ineffective without allowing direct effect. Another way of looking at it is to say that, there would be no purpose to this type of legislation if Member States were allowed to plead their own failure to implement as the reason for the right not being enforced.

As the Court also observed in Van Duyn (1974):

On this basis, while the Court gave little other reasoning it identified that a Directive can indeed be enforced by the means of direct effect provided that the remaining criteria from Van Gend en Loos (1963) are met.

Nevertheless, it must be remembered that, because Directives include objectives that are left for Member States to enact in their own chosen method, if the Member States carry out their obligations, the rights identifiable in the Directives would not be enforceable through EU law but rather through the national law as enacted. For this reason, the direct effect of Directives generally becomes a problem where they are unimplemented or improperly implemented.

Of course, there is nothing to stop the ECJ or the national courts from referring to the Directive after its implementation into national law, in order to ensure that its objectives have indeed been achieved. This was confirmed in the case of Verbond van Nederlandse Ondernemingen v Inspecteur der Invoerrechten en Accijzen (Case 51/76) [1977] ECR 113.

However, it is equally true that during the implementation period the rights contained in the Directive are not enforceable by any means. This would clearly be unfair since the Member States have been given a time limit within which to comply. For this reason, the Court of Justice added an extra criterion to the Van Gend en Loos (1963) test in respect of Directives. In Pubblico Ministero v Ratti (1979) the ECJ stated that direct effect of a Directive is not in question until such time as the implementation period has expired.

Another significant point is that a Directive can only ever be vertically directly effective, given that it is subject to domestic implementation, it could not be granted horizontal direct effect. This means that, according to the reasoning in Van Duyn (1974), an unimplemented or improperly implemented Directive can only be relied upon and enforced against the state, to prevent the latter to plead its own failure to implement as a means to avoid their obligation owed under the Directive.

The ECJ subsequently identified that the national courts should decide against what bodies a Directive could be enforced using vertical direct effect. It has explained also that vertical direct effect may affect not only the state itself but also bodies that could be described as an ‘emanation of the state’ (or ‘arm of the state’). The Court also has devised the test of which bodies can be classed as emanations of the state.

In paragraph 20, the Court then ruled on whether British Gas at the material time (before it was privatised) was such a body. The Court noted it was definitely an ‘emanation of the state’ because it provided a public service, was under the control of the state and was able to exercise special powers. However, when the case returned to the House of Lords, the Law Lords adopted paragraph 20 as the test for determining an ‘emanation of the state’, rather than paragraph 18. According to this test, a body must satisfy three cumulative criteria. It must:

  • be one that provides a public service; and
  • be under the control of the state; and
  • be able to exercise special powers that would not be available to a private body.

It is evident that the paragraph 20 test is much more difficult to satisfy than the paragraph 18 test. Nevertheless, a number of subsequent British cases considered and applied the three-part paragraph 20 test. In Doughty v Rolls Royce plc [1992] 1 CMLR 1045 a publicly owned manufacturing company was held not to be an emanation of the state since it failed the first and third criteria. On the other hand, in Griffin v South West Water [1995] IRLR 15 the national court considered that a privatised water company was an emanation of the state. While the body itself was not as such under the control of the state, certain of the services it operated were.

There is inconsistent application of the principles regarding when a body can be classed as an emanation of the state. Nevertheless, there have been quite liberal interpretations by the ECJ, so that a private company has been held to be an emanation of the state where it is carrying out a public duty.

Where a body is deemed to be an emanation of the state according to the Foster criteria, that body becomes an agent of the state and must therefore take all necessary measures to fulfil the aims prescribed by the EU Directive. In other words, the state as well as its various emanations are all subject to the duty to comply with their obligations under EU law (Portgás (Case C–425/12) [2014] 2 CMLR 30).

Therefore, clarification on whether the Foster test should be based on the judgment paragraph 18 or paragraph 20 was required and provided by the CJEU in the Farrell 2 case (C–413/15 Elaine Farrell v Alan Whitty and Others, EU:C:2017:745). In this case, the CJEU stated that it ‘was not attempting to formulate a general test designed to cover all situations in which a body might be one against which the provisions of a directive capable of having direct effect might be relied upon’. Instead, in Farrell 2 the CJEU confirms that the criteria are not cumulative but alternative, i.e. for a body to be an emanation of the state, it must fulfil one of these three criteria: it is an element of the state and governed by public law; or subject to control by the state or public body; or has been granted special powers to perform a public service.

The fact that Directives may generally only have vertical direct effect inevitably creates major anomalies and injustices where an applicant’s case is against another individual or a private body. This can be seen in the contrasting decisions above and also in Duke v GEC Reliance [1988] 2 WLR 359. The case involved the identical point to that in Marshall (1986), but the employer was not the state but a private company. The House of Lords held that it was not bound to apply Directive 76/207 (now the Recast Directive) because the Directive could not be effective horizontally. Even though the UK was at fault for failing to implement the Directive fully, the availability of a remedy then was entirely dependent on the identity of the employer.

Clearly, the process of ensuring vertical effect of Directives in actions against the state or an emanation of the state was one way of getting round the defects in the definition of ‘Directives’ in Article 288. Nevertheless, the absence of horizontal effect still meant that there were problems in enforcing rights granted by Directives.

In Marshall the Court held that an individual may rely upon a directive against the State, regardless of whether the State is acting as a public authority or employer. The consequence of this ruling is that, for example, a private employee may not rely on a directive but a state employee may. The ruling has been attacked as provoking unjust and anomalous situations, particularly in the field of labour law, where the scope of the Equal Treatment Directive has been reduced. The decision also led to the necessity for an impossibly rigorous definition of the state which proved very difficult to apply in case law.

S Douglas-Scott, Constitutional Law of the European Union (Longman, 2002), p. 296

Nevertheless, the ECJ will pursue its function to ensure that EU law is observed, which implies protecting the doctrine of primacy from any adverse interference by domestic courts which may undermine the effectiveness of the EU legal order should a contradicting national legislation be allowed to trump EU legislation.

Following Mangold v Helm (Case C–144/04) [2005] ECR I–9981, the ECJ suggests that it is possible for a Directive which contains a general principle of EU law (such as non-discrimination based on age) to have horizontal direct effect. Essentially, the ECJ in this case does not allow the horizontal application of the Directive per se, but focuses on ensuring the horizontal application of the fundamental right protected under the General Principles of EU Law (GPEUL), and by extension to the EU Charter of Fundamental Rights.

The means for determining the direct effect of a Directive can be represented in diagram form as shown in Figure 8.2.

A flow diagram depicts the requirements to prove the vertical direct effect of a directive with “yes” or “no” conditions.

Long Description for Figure 8.2

The flow begins with the question, “Does the measure satisfy the Van Gend en Loos criteria other than non-dependency?”. If ‘yes,’ then “Does the measure satisfy the Ratti (1979) criterion?”. If ‘yes,’ then “Is the action against the state, or an emanation of the state, satisfy the Foster (1990) criteria?”. If ‘yes,’ then “The measure is vertically and directly effective against that body”. All the ‘no’ conditions lead to “The measure has no vertical direct effect”.

Figure 8.2 The requirements to prove vertical direct effect of a Directive

Where a Directive does not fulfil the Van Gend en Loos (1963) criteria for direct effect, the ECJ will nevertheless endeavour to ensure that EU law is still applied through the principle of consistent interpretation, also known as indirect effect.

8.4.2 Indirect effect/consistent interpretation

Vertical direct effect is one way in which the ECJ is able to ensure that rights gained through a Directive could be enforced. Still, it has severe limitations. Another way round the problem was the introduction of the process of sympathetic consistent interpretation, or indirect effect.

The ECJ identified that since Member States had an obligation at that time under Article 4 TEU ‘take all appropriate measures … to ensure fulfilment of the obligations arising out of this Treaty’, then the national courts also had a duty to interpret national law so that it would give full effect to EU law (irrespective of whether or not it is directly effective).

This is the principle of indirect effect, also known as consistent/sympathetic interpretation or the Von Colson principle. It applies to EU law in general, not only to Directives, and can obviously be useful in avoiding the issue of the lack of horizontal effect of Directives. The ECJ explained the principle in two cases that were referred to the Court by the German Labour Court:

Von Colson applied to work for a state body, the prison service, while Harz applied to work for a private company. The ECJ identified that the failure to provide compensation did amount to improper or incomplete implementation. Even so, while a remedy would have been available to Von Colson through vertical direct effect, Harz would have been denied a remedy because of the anomaly resulting from lack of horizontal effect. The ECJ took a different and at the time a very novel approach by using the obligation in Article 10 (now repealed) of the EC Treaty to introduce the principle of indirect effect. The German court was bound to give full effect to the Directive and so must order full compensation in both cases.

‘The ECJ in Von Colson (1984) in this way ignored the problem of horizontal and vertical effect, and in fact ignored direct effect in general so it was a means of overcoming the problems seen above in section 8.4.1. Nevertheless, the judgment did leave ambiguous the question of to which national law the process of indirect effect could actually apply. This then allowed the House of Lords to refuse to apply the principle in Duke (1988) even though it would have been a means of providing a remedy for the applicant.

Another limitation in the case itself was that it also left ambiguous or unexplained just how far national courts should go in order to ensure conformity of their own national law and EU law. However, subsequent development or clarification of the principle may have overcome these problems.

The scope of indirect effect provided for by Marleasing (1990) is potentially very wide, then, and it has the effect of introducing horizontal direct effect but by indirect means, hence the title given to the process. However, there does seem to be a significant difference between the Von Colson (1984) approach, which is to do ‘everything possible’ to achieve conformity, and the original approach in Simmenthal SpA v Amministrazione delle Finanze dello Stato (Case 70/77) [1978] ECR 1453, which was to do ‘everything necessary’. The ECJ has more recently held that the principle of indirect effect extends beyond directives and can apply also to a framework decision (Pupino C–105/03 [2005] 2 CMLR 63).

The ECJ does seem to have linked the concepts of direct effect and indirect effect as different ways of dealing with the same problem. In Johnston v Chief Constable of the RUC (Case 222/84) [1986] ECR 1651 the Court suggested that it was the duty of national courts to interpret national law in conformity with EU law (indirect effect) and only if this was not possible to then enforce EU law in preference to inconsistent national law (direct effect).

The ECJ also confirmed the Marshall jurisprudence in Dominguez (Case C–282/10) [2012] 2 CMLR 14, whereby whilst Directives do not have horizontal effect, it is for the national courts to determine whether a Directive could be used by an individual against a private body, particularly if a sympathetic interpretation of national law is not possible in the circumstance.

In Pfeiffer & Others (Joined Cases C–397–403/01) [2004] ECR I–8835, the ECJ further added that when interpreting national law in light of EU law, domestic courts should take the whole body of national legislation into consideration and not just the implementing measure – so that domestic courts may select the relevant national legislation to interpret in a way that is consistent with the EU Directive and facilitates the fulfilment of its objectives.

There are limitations to the process. In Arcaro (Case C–168/95) [1996] ECR I–4705 the ECJ held that the provisions of an unimplemented Directive cannot be imposed on an individual where to do so would lead to criminal liability. The most important limitation, of course, is that the process is entirely dependent on the willingness of the national courts to use it. As has been seen in Duke (1988), the national courts are not always so willing so that there is the possibility of lack of uniformity throughout the EU. Inconsistency indeed has been shown. The House of Lords in Litster v Forth Dry Dock and Engineering Co [1989] 1 All ER 1134 had to consider the incompatibility of the then English TUPE law and provisions of the Acquired Rights Directive 77/187. The Court, taking a different approach from that in Duke (1988), felt bound to interpret the English Regulations in a way that would give full effect to the Directive.

the indirect application of EC directives by national courts cannot be guaranteed. This reluctance on the part of the national courts to comply with the Von Colson principle, particularly as applied in Marleasing, is hardly surprising. It may be argued that in extending the principle of indirect effect in this way the ECJ is attempting to give horizontal effect to directives by the back door, and impose obligations, addressed to Member States, on private parties, contrary to their understanding of domestic law. Where such is the case, as the House of Lords remarked in Duke this could be most unfair.

J Steiner and L Woods, Textbook on EC Law (8th edn, Oxford University Press, 2003), p. 109

8.4.3 ‘Incidental’ horizontal effect

In Faccini Dori v Recreb Srl (Case C–91/92) [1994] ECR I–3325 Advocate-General Lenz argued that Directives should be capable of horizontal direct effect because of changes in the TEU and also because the fact that they were not meant that individuals were being denied rights that they otherwise might claim legitimately through EU law. However, the ECJ did not follow his argument but instead applied the Marshall (1986) criteria in the case.

Three arguments were originally used to deny the horizontal effect of Directives:

  • The lack of legal certainty – however, now that all Directives are published, the argument cannot apply and it is in fact the lack of horizontal effect and now the main uncertainties are what bodies will be considered public for vertical effect and the circumstances in which national courts will be prepared to impose indirect effect.
  • The estoppel argument, that since individuals had no control over the implementation of Directives they were not at fault and should not be fixed with fault – however, the widening of the concept of the state in Foster in effect means that that is precisely what is happening.
  • That to allow horizontal effect to Directives was to blur the distinction between Regulations and Directives – this again is not really an argument since, if Member States implemented Directives as national law, there would effectively be no distinction anyway.

One interesting offshoot of the problems associated with direct effect of Directives has arisen in situations where individuals have tried to exploit the principle of direct effect to establish the illegal nature of the national law rather than to try to enforce any EU rights. The way that such situations have been dealt with has become known as ‘incidental’ or ‘triangular’ horizontal effect. Some recent cases stand out, and they appear to go against the eventual stance taken in Dori and seem to indicate a way in which Directives can indeed be horizontally effective.

The first allowed a party to use a Directive in effect horizontally in a defence against another private party where that party was relying on national law which had not fully followed the Directive.

A similar result has been achieved in a contractual dispute where the Directive was merely incidental to proving the breach of contract and had nothing to do with enforcing an EU right.

The above cases relate to the ‘incidental’ effect of Directives whereby national law is disapplied due to a failure by the Member States to meet the measure’s procedural or technical requirements – there is creation or application of individual rights through the Directive.

By contrast, the ‘triangular’ horizontal effect of a Directive implies that a Directive is given effect against the state which has an incidental effect on a third party. In Case C–201/02 Wells a company had been granted a mining permission to carry out work in a quarry near the claimant’s dwelling. However, contrary to Directive 85/337, no environmental impact assessment (EIA) had been performed prior to the permit being granted. The ECJ allowed the claimant to invoke the Directive against the UK Secretary of State for Transport for its failure carry out an EIA. Whilst the obligation under the Directive was imposed on the UK, the effect of this judgment is that the mining permit could be suspended or revoked. This jurisprudence was confirmed in Arcor (Cases C–152 and 154/07) [2008] ECR I–5959.

A flow diagram depicts the possible means of enforcing rights contained in Directives with “yes” or “no” conditions.

Long Description for Figure 8.3

The flow begins with the question, “Is the directive clear and precisely stated?”. If ‘no,’ then “there are no enforceable rights,” and if ‘yes,’ then “Has the date for implementation passed?” If ‘no,’ then “There is no European Union law to enforce and no national law yet in place to enforce,” and if ‘yes,’ then “Is the body against whom the claim is made the state? Or is the body an emanation of the state?”. If ‘yes,’ then “The directive can be enforced under vertical direct effect” and if ‘no,’ then “Can the national law be interpreted harmoniously so as to give effect to the directive?”. If ‘yes,’ then “The national court may enforce the measure, but the court may see this as unfair against a private party” and if ‘no,’ then “Are Brasserie du Pecheur criteria satisfied?”. If ‘yes,’ then “Claimant may use the state for damage state liability” and if ‘no,’ then “No liability”.

Figure 8.3 The possible means of enforcing rights contained in Directives

SUMMARY

  • Direct effect, indirect effect/consistent interpretation and state liability are all devices created by the Court of Justice to ensure effective enforcement of the Treaties.
  • Direct effect means the EU measure can be enforced before domestic courts, and the Court of Justice has established a test of direct effect – the measure must be clear, precise and unconditional.
  • Direct effect can be both horizontal (enforceable against other citizens of the EU) and/or vertical (enforceable against the state or an emanation of the state).
  • Complex issues arise in the case of unimplemented directives, as directives are conditional upon domestic implementation by Member States – so the Court of Justice also created indirect effect/consistent interpretation, a means by which national courts are empowered to interpret national legislation in light of EU law provision, and in line with the Member States’ EU obligations, in order to uphold the effectiveness, usefulness and primacy of EU law.

Further reading

Articles

  • Albors-Llorens, A, ‘The Direct Effect of EU Directives: Fresh Controversy or a Storm in a Teacup? Comment on Portgás’ (2014) 39 EL Rev 851.
  • Bakhtina, KR, ‘Farrell II and the Concept of “Emanation of the State”’ (2018) 18(1) ICLR.
  • Ciacchi, AC, ‘The Horizontal Effect of EU Fundamental Rights’ (2019) 15 ECL Rev 294.
  • Dougan, M, ‘When Worlds Collide: Competing Visions of the Relationship Between Direct Effect and Supremacy’ (2007) 44 CML Rev 931.
  • Drake, S, ‘Twenty Years After Von Colson: The Impact of “Indirect Effect” on the Protection of the Individual’s Community Rights’ (2005) 30 EL Rev 329.
  • Frantziou, E, ‘(Most of) the Charter of Fundamental Rights is Horizontally Applicable’ (2019) 15 ECL Rev 306.
  • Mancini, F, ‘The Making of a Constitution for Europe’ (1989) 26 CML Rev 595.
  • Prechal, S, ‘Does Direct Effect Matter?’ (2000) 37 CML Rev 1047.
  • Prechal, S, ‘Horizontal Direct Effect of the Charter of Fundamental Rights of the EU’ (2020) 66 Rev Derecho Comunitario Eur 407–426.

Books

  • Craig, P and de Burca, G, EU Law: Text, Cases and Materials (OUP, 2020).
  • Schütze, R, European Union Law (OUP, 2021).

9 State liability

DOI: 10.4324/9781003218562-9

9.1 Introduction

As we have seen in Chapter 8, both direct effect (in particular, the lack of horizontal direct effect of Directives) and indirect effect have limitations and leave situations where a party could be without a remedy because of the failure of a Member State to implement a Directive or because of improper implementation. The ECJ devised a third way of avoiding the problems identified in Chapter 8 – this is state liability.

State liability for breaches of EU law was introduced in Francovich & Bonifaci v Italy (Cases C–6/90 and C–9/90) [1991] ECR I–5357, when the ECJ declared that the Member States were ‘obliged’ to compensate individuals for loss and damage caused to them by breaches of EU law. The Court stated:

There is nothing explicitly in either of the Treaties to support the imposition of liability on Member States, but the Court justified its decision by finding that it was stated implicitly, in what is now Article 4(3) of the TEU, according to which Member States have an obligation to ensure the fulfilment of their obligations arising under the Treaties.

9.2 The conditions for state liability

Francovich & Bonifaci (1991) was concerned specifically with the failure of the Italian government to implement a Directive. The obligation imposed on states to implement Directives is clear (see Article 4(3) TEU and Article 288 TFEU). However, two issues remained unclear:

  • what the position might be in relation to other breaches; and
  • the level of fault required to establish liability.

The ECJ addressed these issues in Brasserie du Pêcheur SA v Germany; Factortame III (Cases C–46 and 48/93) [1996] ECR I–1029. The Court made it clear that state liability was a general principle, not restricted to a failure to implement Directives, and that three conditions had to be satisfied:

  1. the rule of EU law infringed must have been intended to confer rights on individuals;
  2. the breach must be sufficiently serious; and
  3. there must be a direct causal link between the breach of the obligation resting on the state and the damage sustained by the injured parties.

The burden of proof is on the claimant to establish all three conditions. State liability claims are brought in the national court of the defendant state.

9.2.1 Condition (1): An intention to confer rights on individuals

There is a strong link here to direct effect. The case of Köbler v Austria (Case C–224/01) [2003] ECR I–10239, concerned an alleged breach of Article 45(2) TFEU (the right of migrant workers to work without discrimination on grounds of nationality – see Chapter 12). As A-G Léger noted in that case: ‘The rule of law purportedly infringed … is directly effective and its purpose is therefore necessarily to confer rights on individuals’ (emphasis added). Condition (1) has been held to be satisfied in the case of several directly effective Treaty articles, including Articles 34 and 35 TFEU (the free movement of goods – see Chapter 14) and Articles 49 and 56 TFEU (the right of companies and self-employed people to ‘establish’ themselves in another Member State, or to provide services to someone from another Member State – see Chapter 13). A table at the end of this chapter summarises some of the provisions of EU law which have been held to satisfy the first condition of state liability.

If the first condition is satisfied, the focus shifts to the second condition; but if not then the claim fails. This happened in Paul & Others v Germany (Case C–222/02) [2004] ECR I–9425, where the ECJ decided that Article 3(1) of Directive 94/19 was not intended to confer rights on individuals. In Berlington & Others v Hungary (Case C–98/14) [2015] 3 CMLR 45, the ECJ held that whilst Article 56 TFEU (the freedom to provide services) was intended to confer rights on individuals, Directive 98/34 did not do so.

9.2.2 Condition (2): A sufficiently serious breach

The test for determining the second condition, a ‘sufficiently serious’ breach, is whether the Member State ‘manifestly and gravely disregarded the limits on its discretion’ (Brasserie; Factortame III (1996)). In two situations, a breach will be deemed automatically to have been serious. The first situation is when a Member State fails to take any measures to implement a Directive. In Dillenkofer & Others v Germany (Case C–178/94) [1996] ECR I–4845, the Court stated (emphasis added):

The second situation is when the Member State defies clear ECJ case law. For example, in Fuß v Stadt Halle (Case C–429/09) [2010] ECR I–12167, the ECJ pointed out that employment conditions imposed on fire-fighters in the German city of Halle were not only contrary to Directive 2003/88 (the Working Time Directive – see Chapter 17) but at least three earlier ECJ decisions. Unsurprisingly, this was held to be a serious breach of EU law. The Court stated that (emphasis added):

In principle it is for the national courts to decide whether or not a breach of EU law was ‘sufficiently serious’. However, the ECJ has identified various factors which may be used, as follows:

  • the clarity and precision of the rule breached;
  • the measure of discretion left by that rule;
  • whether the infringement and the damage caused was intentional or involuntary;
  • if there was an error of law, whether the error was excusable or not;
  • whether the position taken by one of the Union’s institutions may have contributed towards the adoption or retention of national measures or practices contrary to EU law.

Clarity and precision

Many state liability cases involve the failure to properly implement a Directive. Such breaches may or may not be sufficiently serious; it often depends on the clarity of the wording in the Directive. In BT (Case C–392/93) [1996] ECR I–1631 and Denkavit International v Germany (Case C–283/94) [1996] ECR I–5063, the failures by (respectively) the UK and Germany were held not to be serious, largely because the text of the Directives lacked clarity or precision. Moreover, both states had acted in good faith in trying to implement the Directive correctly (suggesting the breach was also excusable).

Conversely, in Rechberger & Others v Austria (Case C–140/97) [1999] ECR I–3499 and Stockholm Lindöpark v Sweden (Case C–150/99) [2001] ECR I–493, incorrect implementation was deemed to be serious, because in both cases the relevant Directives were clearly worded (also suggesting that the breach was inexcusable).

Discretion

Where a Member State has broad discretion under EU law, it will be more difficult to prove that a breach has occurred at all, let alone that it was a serious breach. For example, in Schmidberger v Austria (Case C–112/00) [2003] ECR I–5659, it was alleged by the claimant, a German road haulage firm, that the Austrian government had seriously breached Article 34 and/or 35 TFEU by authorising a road closure in order to allow an environmental demonstration to take place. The ECJ rejected the allegation, noting that the Austrian government had to be given a ‘wide margin of discretion’, given the need to balance competing (if not contradictory) interests: the need to protect the fundamental rights of the protestors to exercise their right to free speech (on one hand) and the economic rights of the claimants to import and export goods across international borders (on the other). The Court held (emphasis added):

Conversely, a lack of discretion is much more likely to lead to a conclusion that the breach was sufficiently serious. For example, the breach of Article 35 TFEU by the UK in Hedley Lomas (Case C–5/94) [1996] ECR I–2553 was deemed to be serious because of the lack of discretion available. In that case, the Court stated (emphasis added):

A similar outcome was seen in Rechberger & Others v Austria (1999). In this case, the Austrian government had unilaterally postponed the implementation date of Directive 90/134 (the Package Travel Directive), specifically Article 7 of it, something which is not permitted under any circumstances. This allegedly caused the claimants financial loss. The ECJ held that the breach was sufficiently serious, essentially because of the lack of discretion:

There is often a correlation between the clarity of the rule breached and a (lack of) discretion. For example, in Stockholm Lindöpark v Sweden (2001), the rule of EU law breached – the Sixth VAT Directive – was clearly worded (suggesting a serious breach). Moreover, the clarity of the rule breached also meant that the Swedish government lacked discretion (confirming the seriousness of the breach). The Court held that ‘Given the clear wording of the Sixth Directive, [Sweden] was not in a position to make any legislative choices’ (emphasis added).

Intentional breach

The third factor involves asking whether or not the breach was intentional (which makes it more likely to be serious). For example, in Factortame III [1999] 3 WLR 1062, when the case was returned to the House of Lords from the ECJ, the House ultimately held that the breach of Article 49 TFEU by the UK was serious because the government had intentionally breached the provision in question. Lord Slynn stated (emphasis added):

By way of contrast, in Negassi v Home Secretary [2013] EWCA Civ 151; [2013] 2 CMLR 45, the Court of Appeal held that the UK government’s failure to properly implement Directive 2003/9 was insufficiently serious. One of the reasons for this decision by the Court was the lack of intent on the part of the government. Kay LJ said (emphasis added):

Excusable breach

The fourth factor is whether or not the breach was excusable. In several of the cases discussed above, the Member States had acted in good faith which suggested that any breach was excusable (BT, Denkavit). The breach of Article 34 TFEU by the German government in Brasserie du Pêcheur was ultimately deemed to be insufficiently serious because it was an excusable error (see the decision of the German Federal Court of Justice in Brasserie du Pêcheur v Germany (No.2) [1997] 1 CMLR 971).

Involvement of an EU Institution

This final factor is rarely invoked, but it can be important. In BT (1996), one of the reasons given for rejecting the claim was the involvement (or lack thereof) of the European Commission. The UK government had drafted its implementing legislation which was sent to the Commission, but no response was forthcoming. Had the Commission responded and pointed out the error, the implementing legislation could have been amended and the litigation avoided entirely.

9.2.3 Condition (3): Direct causal link

As to the third condition, the ECJ has held that it is, generally speaking, for the national courts to determine whether a direct causal link between breach and damage exists (Rechberger & Others (1999)). In Leth & Others v Austria (Case C–420/11) [2013] 3 CMLR 2, the Court stated:

9.3 What is the ‘state’?

Most state liability cases involve actions against central government. In the UK, the vast majority of state liability actions were brought against central government, including the cases (mentioned above) of Factortame III, Hedley Lomas and BT. Usually a specific government Minister was named, as in Negassi v Home Secretary [2013] EWCA Civ 151, [2013] 2 CMLR 45; Delaney v Secretary of State for Transport [2014] EWHC 1785, [2014] 3 CMLR 32; and Recall Support Services Ltd & Others v Secretary of State for Culture, Media & Sport [2014] EWCA Civ 1370, [2015] 1 CMLR 38. This is also the case when state liability claims are brought in the Republic of Ireland (see Ogieriakhi v Minister for Justice & Equality & Others (No.2) [2015] 3 CMLR 18).

However, state liability actions can be brought against other bodies. In Berlington & Others (2015), the ECJ held that the ‘principle of Member State liability is applicable, inter alia, where the national legislature was responsible for the infringement’. In Konle v Austria (Case C–302/97) [1999] ECR I–3099 the ECJ decided that compensation need not necessarily be the responsibility of central government, and that state liability claims could instead be brought against regional government. The Court stated:

Long Description for Figure 9.1

The flow begins with the question, “Has European Union rule been breached, which was intended to confer rights?”. If ‘yes,’ then “Is the breach sufficiently serious to justify imposing liability on the state?”. If ‘yes,’ then “Is there a direct causal link between the breach of European Union law by the state and the damage suffered by the claimant?”. If ‘yes,’ then “state liability can be imposed under Francovich (1991) or Brasserie du Pecheur (1996) criteria”. All the ‘no’ conditions lead to “There is no state liability”.

Figure 9.1 The criteria to establish state liability under Brasserie du Pêcheur (1996)

This is very significant as it means claimants can pursue state liability actions against regional or local governments, rather than central governments, when it is the former who are responsible for the breach in question. The proposition that regional or local authorities can be held liable extends beyond countries with a federal structure. For example, in Combinatie & Others (Case C–568/08) [2010] ECR I–12655, a state liability case was brought against the province of Drenthe, in the north-eastern Netherlands, rather than the Dutch government. In Barco De Vapor [2014] EWHC 490 (the facts of which were markedly similar to those in Hedley Lomas (1996)), state liability was imposed on Thanet District Council in Kent. In Haim (Case C–424/97) [2000] ECR I–5123, the ECJ stated that state liability claims could be brought irrespective of the ‘public authority … responsible for the breach’. The Court held that (emphasis added):

This is intriguing but arguably begs more questions than it answers – in particular, what exactly is the meaning and scope to be given to the phrase ‘public authority’ here? In Haim itself, the defendant body was the Association of Dental Practitioners in the Nordrhein region of Germany. This demonstrates that professional, regulatory bodies (at least) are included in the concept of ‘public authority’.

Even breaches of EU law by the judiciary may be lead to the imposition of state liability, according to the ECJ in Köbler v Austria (2003). Although the claim in that case (involving an alleged breach of EU law by the Austrian Supreme Administrative Court) failed, the ECJ expressly ruled that national courts ‘adjudicating at last instance’ could face liability albeit that such a case would have to be ‘exceptional’. The Court stated (emphasis added):

This was confirmed in Traghetti del Mediterraneo v Italy (Case C–173/03) [2006] ECR I–5177, involving a claim against the Italian Supreme Court. However, the ECJ stressed that ‘State liability can be incurred only in the exceptional case where the national court adjudicating at last instance has manifestly infringed the applicable law’. More recently, in Ferreira da Silva e Brito & Others v Portugal (Case C–160/14), [2016] 1 CMLR 26, the ECJ held that the non-referral of a question to the Court of Justice under the preliminary rulings procedure by the Portuguese Supreme Court could attract state liability. The Court of Justice pointed out that, in cases such as the present case, which was ‘char-acterised both by conflicting lines of case-law at national level’ and by the fact that the provision of EU legislation in question (Directive 2001/23) ‘frequently gives rise to difficulties of interpretation in the various Member States’, a national court of last resort ‘must comply with its obligation to make a reference to the Court’.

This development had been predicted by Joxerramon Bengoetxea, ‘Text and Telos in the European Court of Justice’ (2015) 11 ECL Rev 184, who wrote that ‘I would not exclude the possibility of the Köbler (or Traghetti) jurisprudence being applied to a stubborn national court of last instance abusively declaring acte clair in situations that should have been submitted for preliminary ruling’.

Finally, in AGM v Finland (Case C–470/03) [2007] ECR I–2749 the ECJ introduced a form of vicarious liability when it stated that EU law ‘does not preclude an individual other than a Member State from being held liable, in addition to the Member State itself, for damage caused to individuals by measures which that individual has taken’ in breach of EU law. Thus liability could, in principle, be imposed on a Member State for breaches of EU law made by individual government employees.

9.4 Limitations on the recovery of compensation

Member States are allowed to impose conditions on the amount of damages that the claimant can recover in state liability claims. However, any such conditions must not be less favourable than those relating to similar domestic claims (the ‘principle of equivalence’) or framed in such a way as to make it in practice ‘impossible or excessively difficult’ to obtain reparation (the ‘principle of effectiveness’). These principles were held to be applicable in state liability cases by the ECJ in Brasserie/Factortame III (1996) in which the Court stated:

9.4.1 The principle of equivalence

This ‘principle’ means that national rules on the recovery of damages via civil litigation are deemed to be also applicable in state liability actions. One question that arose in the Factortame III case was whether ‘exemplary’ (or ‘punitive’) damages were available in state liability. The ECJ answered yes, provided that such damages were available in the Member State in domestic civil litigation cases. The Court stated:

This was applied by the High Court in Santos v Home Secretary [2016] EWHC 609. Having found that the defendant, the UK Home Secretary (which at the time of the case was the UK’s future Prime Minister, Theresa May), had committed a serious breach of EU law, Lang J had to decide whether to award exemplary damages. He said:

Another example of the principle of equivalence is limitation periods (a period of time during which a claimant is required to bring civil litigation, after which their case becomes time-barred). National limitation periods are applied in state liability cases – subject to one proviso: they must be ‘reasonable’. In Danske Slagterier v Germany (Case C–445/06) [2009] ECR I–2119, the ECJ stated that reasonable time limits for bringing proceedings were justifiable ‘in the interests of legal certainty which protects both the taxpayer and the authorities concerned’.

9.4.2 The principle of effectiveness

This ‘principle’ means that any procedural rule of domestic civil litigation is deemed to be inapplicable in state liability cases where it would make the recovery of damages ‘impossible or excessively difficult’. Another issue in the Factortame III case was whether damages for ‘pure’ economic loss were recoverable against the UK. In domestic English civil litigation, there is a rule that such damages are not normally available. The principle of equivalence (above) states that this rule should be applied in state liability cases against the UK. However, doing so would render it practically ‘impossible’ for claimants in such cases – most of whom are seeking ‘pure’ economic loss, typically loss of profits – to get any compensation. Therefore, the principle of effectiveness applies here to render the English rule inapplicable. In Brasserie/Factortame III (1996), the Court stated:

Another potential obstacle to the recovery of damages in state liability claims is the rule of ‘res judicata’ (literally: let the judgment rest), which is a typical feature of domestic civil litigation. In essence, it means that once a case has been litigated, and judgment given, the issue should not be endlessly re-litigated. This is important in the interests of legal certainty, but is potentially problematic if a claimant in a state liability action has already brought an action in domestic civil litigation (and hence there is now a judgment on the matter). Is it possible to bring a state liability action despite res judicata? The answer is yes – because of the principle of effectiveness. In Ferreira da Silva e Brito & Others v Portugal (2016), the Court stated (emphasis added): The following is a table of EU law provisions found to satisfy the first condition of State Liability: an intention to confer rights on individuals.

EU law provision Right conferred on individuals Case(s)
Article 34 TFEU To import goods from other Member States Brasserie du Pêcheur v Germany (1996); Schmidberger v Austria (2003); Danske Slagterier v Germany (2009)
Article 35 TFEU To export goods to other Member States Hedley Lomas (1996); Barco De Vapor (2014)
Article 45(2) TFEU To work in another Member State without discrimination on grounds of nationality Köbler v Austria (2003)
Article 49 TFEU To establish a business in another Member State Factortame III (1996); Haim (2000)
Article 56 TFEU To provide services in another Member State Berlington & Others v Hungary (2015)
Article 63 TFEU The free movement of capital Konlev Austria (1999)
Directive 85/337 To have the environmental effects of a project assessed by the competent authorities Leth v Austria (2013)
Article 7 of Directive 90/314 To a refund and/or repatriation in the event of a tour operator&s insolvency Dillenkofer & Others v Germany (1996); Rechberger & Others v Austria (1999)
Directive 98/37 To import and export machinery AGM v Finland (2007)
Article 2(1) of Directive 2000/78 To work without discrimination on grounds of age Specht & Others (2015)
Directive 2001/23 The safeguarding of employees’ rights in the event of a transfer of undertakings Ferreira da Silva e Brito & Others v Portugal (2016)
Article 6(b) of Directive 2003/88 To work for no more than 48 hours per week on average FußvStadt Halle (2010)
Articles 9 and 10 of Directive 2004/38 The entitlement of migrant citizens and their family members to Residence Cards Santos v Home Secretary (2016)
Article 16(2) of Directive 2004/38 To work in another EU Member State (as the family member of an EU citizen) Ogieriakhi v Minister for Justice & Equality & Others (No. 2) (2015)
Articles 13 and 18 of Directive 2004/83 To the grant of refugee or ‘subsidiary protection’ status D v Home Secretary (2015)
Article 12(2) of Directive 2004/80 Payment of ‘fair and appropriate compensation’ to victims of violent crime B.V. v Italy (Case C-129/19)

In Târṣia v Romania (Case C–69/14) [2016] 1 CMLR 47, the ECJ held that where it was discovered that a national supreme court had reached an incorrect decision on a point of EU law (and potentially committed a sufficiently serious breach of EU law), but where it had subsequently become impossible to re-litigate the point in the domestic courts because of res judicata, ‘individuals cannot be deprived of the possibility of rendering the State liable in order to obtain legal protection of their rights’.

SUMMARY

  • In Francovich, the ECJ created the possibility of individuals seeking compensation against a state whose breach of EU law caused that individual to suffer loss. State liability is based on Article 4(3) TEU.
  • State liability is not limited to the failure by the state to implement a directive, as in Francovich, although that still remains a good example (Dillenkofer v Germany, BT, Rechberger & Others, Stockholm Lindöpark). State liability extends to breaches of EU law by Member States generally.
  • In Brasserie du Pêcheur/Factortame III, the ECJ identified three conditions for state liability: (1) the rule of EU law infringed must be intended to confer rights on individuals; (2) the breach of EU law must be ‘sufficiently serious’; and (3) there must be a direct causal link between the state’s failure to comply with EU law and the damage suffered by the individual.
  • There is a strong correlation between condition (1) and direct effect. Most state liability cases (other than those involving a failure to implement a directive) involve breaches of directly effective provisions of the TFEU. Examples include Brasserie du Pêcheur (Article 34 TFEU), Hedley Lomas (Article 35 TFEU), Köbler (Article 45(2) TFEU), Factortame III (Article 49 TFEU) and Berlington & Others (Article 56 TFEU).
  • For condition (2), a ‘sufficiently serious’ breach, the decisive test is whether there had been a ‘manifest and grave’ disregard of the limits of the state’s discretion. The ECJ has identified several ‘factors’ in seeking to prove this:
    • the clarity and precision of the rule breached,
    • the measure of discretion left by that rule to the national authorities,
    • whether the infringement was intentional or involuntary,
    • whether any error of law was excusable or inexcusable,
    • the fact that the position taken by a Union institution may have contributed towards the adoption or maintenance of national measures or practices contrary to EU law.
  • Two breaches of EU law are automatically ‘sufficiently serious’:
    • Failure to implement a Directive on time (Dillenkofer & Others).
    • Defiance of clear ECJ case law (Fuß v City of Halle).
  • For condition (3), it is for the national courts to determine whether the causal link exists (Brasserie du Pêcheur/Factortame III; Leth).
  • Damages for state liability are available regardless of the origin of the state responsible for the breach. Most state liability cases involve central government (Francovich, Brasserie du Pêcheur, Factortame III). However, it is possible to bring a state liability case against the national legislature (Berlington & Others); a regional government (Konle, Fuß); a public body, such as a professional regulatory body (Haim).
  • National courts adjudicating at last instance may also be held liable, albeit exceptionally (Köbler, Traghetti del Mediterraneo). Failing to seek a preliminary ruling when the point of EU law is not acte clair could lead to state liability (Ferreira da Silva e Brito & Others).
  • An individual could be held liable under state liability in addition to the state itself, a form of vicarious liability (AGM).
  • Reparation is subject to two principles (Brasserie du Pêcheur, Factortame III):
    • The principle of equivalence. Reparation must be made in accordance with national rules on civil liability, including entitlement to exemplary damages and limitation periods, but under conditions that are no less favourable than for similar domestic claims.
    • The principle of effectiveness. National rules on recovery of compensation in state liability claims cannot be so framed so as to make it in practice ‘impossible or excessively difficult’ to obtain compensation. Thus, total exclusion of loss of profit in the context of economic and commercial litigation is not acceptable as far as state liability claims are concerned.

Further reading

Articles

  • Albors-Llorens, A, ‘The Principle of State Liability in EC Law and the Supreme Courts of the Member States’ (2007) 66 CLJ 270.
  • Anagnostaras, G, ‘Not as Unproblematic as You Might Think: The Establishment of Causation in Governmental Liability Actions’ (2002) 27 EL Rev 663.
  • Bengoetxea, J, ‘Text and Telos in the European Court of Justice’ (2015) 11 ECL Rev 184.
  • Breuer, M, ‘State Liability for Judicial Wrongs and Community Law: The Case of Köbler v Austria’ (2004) 29 EL Rev 243.
  • Davis, RW, ‘Liability in Damages for a Breach of Community Law: Some Reflections on the Question of Who to Sue and the Concept of “the State”’ (2006) 31 EL Rev 69.
  • Lock, T, ‘Is Private Enforcement of EU Law through State Liability a Myth? An Assessment 20 Years after Francovich’ (2012) 49 CML Rev 1675.
  • Nassimpian, D, ‘… And We Keep on Meeting: (De)fragmenting State Liability’ (2007) 32 EL Rev 819.
  • Prechal, S, ‘Member State Liability and Direct Effect: What’s the Difference After All?’ (2006) 17 EBL Rev 299.

10 The Internal Market

DOI: 10.4324/9781003218562-10

10.1 The aims of the Internal Market

The Internal Market principle is located in Article 26(2) TFEU. It states:

Following the EU’s enlargements in 2004, 2007 and 2013, the Internal Market was – in population terms at least – the world’s largest single trading bloc (overtaking NAFTA [the North American Free Trade Association], comprising the USA, Canada and Mexico). Post-Brexit, NAFTA has retaken pole position (with a combined population of nearly 500 million compared to the EU’s total population of nearly 450 million). However, the Internal Market is much more than a free trade area (which characteristically only seeks to remove trade barriers between its Member States). There is also a common policy on the imposition of Customs duties both internally and externally (known as a Customs union), and a whole series of rules promoting the free movement of persons, services and capital as well as goods. NAFTA is essentially a free trade area, as is EFTA (the European Free Trade Association) which the UK helped form in the late 1950s as a ‘rival’ organisation to what was then the EEC. EFTA still exists, although the only surviving members are Iceland, Norway and Liechtenstein (all of whom are members of the European Economic Area [EEA]) and Switzerland. The purpose of the EEA was to allow those European states who were not committed to full EU membership to obtain some of the benefits of the Internal Market.

The Internal Market is the cornerstone of the EU. Free movement in EU law has several mechanisms and produces numerous advantages:

  • Removal of trade barriers ensures that manufacturers and producers will have easier access to more consumers.
  • As a corollary, those consumers will have access to more, and better, goods.
  • Economies of scale and increased competition from manufacturers and producers in other countries will also drive down the prices of goods.
  • The abolition of Customs charges and other tariffs will stop individual states from protecting inefficient domestic manufacturers and producers.
  • Removal of obstacles to personal movement helps workers (and potential workers) to move from areas of high unemployment and/or areas of low wages to areas of high employment and/or better wages in different countries. Mutual recognition of vocational qualifications and practical experience obtained in different countries facilitates the movement of workers and self-employed people.
  • The abolition of discrimination based on nationality (except where absolutely necessary) ensures employers taking on those foreign workers cannot exploit them vis-à-vis domestic workers, further encouraging free movement.

10.2 The ‘Four Freedoms’

The ‘Four Freedoms’ are:

  1. the free movement of persons;
  2. the free movement of services;
  3. the free movement of goods; and
  4. the free movement of capital.

The first three of these will be explored in detail in Chapters 11 and 12 (persons), 13 (services), and 14 and 15 (goods). This chapter will look at some of the issues that have arisen that are common to all Four Freedoms, and those issues which involve the relationship between one or more of them. The fourth freedom, capital, will also be discussed briefly.

10.2.1 Common features

The free movements of persons, services, goods and capital all involve essentially the same basic principle: the removal of all barriers to movement imposed by national legislation, regulation or administration except those which are objectively necessary in order to protect some essential national interest. In other words, there is not ‘absolute’ free movement. Member States of the EU are entitled to retain in force, or even introduce new, barriers to free movement. However, they must be able to justify those barriers, either by invoking specific derogations contained in the TFEU or by invoking derogations authorised by the ECJ.

The TFEU allows Member States to derogate from the free movement of persons, services, goods and capital on the grounds of public policy and public security. These are the only derogations common to all Four Freedoms. There are other derogations which are specific to one or more of the free movements. Derogations on grounds of ‘public health’ are available for the freedoms of persons and services, while Article 36 allows Member States to derogate from the free movement of goods on the ground of ‘human health protection’. Article 36 also allows Member States to derogate from the free movement of goods on grounds of ‘public morality’, ‘the protection of national treasures possessing artistic, historic or archaeological value’ and ‘the protection of industrial or commercial property’. None of these derogations applies to the other freedoms.

In addition to the specific Treaty derogations, the ECJ has introduced and developed a doctrine under which some restrictions on the freedoms imposed by national legislation, regulation or administration can be justified on any other grounds (sometimes referred to as a ‘rule of reason’). However, four criteria have to be satisfied. In Gebhard (Case C–55/94) [1995] ECR I–4165, the ECJ stated that, in order for a national rule which restricts any of the freedoms to be justified, it must:

  1. be non-discriminatory;
  2. be justified by imperative requirements in the general interest;
  3. be suitable for the attainment of the objective which it pursues;
  4. not go beyond what is necessary in order to attain its objective (the ‘proportionality’ doctrine).

The application of these principles to specific cases will be dealt with in the respective chapters (and later in this chapter). However, it should be noted that the ECJ’s application of a uniform, four-part test to the three freedoms of movement for persons, services and goods has attracted academic criticism. Luigi Daniele has argued that there is a lack of common ground between the movement of persons and the movement of services. The former involves the physical movement of a person to another EU Member State on a more or less permanent basis; the latter requires only that a service is provided by a person established in one Member State and received by a person from another Member State. There is no requirement of physical movement at all. Daniele suggests that applying the same rules to fundamentally different situations can create more problems than it solves. He concludes by suggesting:

The global approach to the free movement of persons risks bringing the Court unduly to put on the same footing rather different situations. More care should be taken in future when transposing the results achieved by case law in one field to another…. The Court should refrain from relying too much on the so-called ‘rules of reason’. The uncertainty of the criteria on which such rules are based often forces the Court to follow a case-by-case approach. It is therefore extremely hard for the interpreter to predict how the Court will decide a given case. This might also cause an undesirable flow of mostly ill-founded cases. Before the situation gets out of hand, a re-consideration would be welcomed.

L Daniele, ‘Non-Discriminatory Restrictions to the Free Movement of Persons’ (1997) 22 EL Rev 191

As of 2023, however, the ECJ shows no sign of abandoning its ‘globalisation’ policy of applying the same ‘rules of reason’ to the free movements of persons, services and goods. Nor has it shown any intention of tightening up the criteria on which its ‘rules’ are based.

10.2.2 The free movement of capital

Space precludes a detailed analysis of the free movement of capital provisions. The key provisions are found in Articles 63–66 TFEU. Article 63(1) provides that ‘within the framework of the provisions set out in this chapter, all restrictions on the movement of capital between Member States and between Member States and third countries shall be prohibited’.

The original Treaty provisions on the free movement of capital were far less sweeping, however, and liberalisation of capital movements has been a slower process than that of the other freedoms. In Casati (Case 203/80) [1981] ECR 2595, for example, the ECJ stated that complete freedom of movement of capital could undermine the economic policy of one of the Member States or create an imbalance in its balance of payments. According to the ECJ in Bordessa and Others (Cases C–358 and 416/93) [1995] ECR I–361, full liberalisation of capital movements did not occur until the Council adopted Directive 88/361.

‘Capital’

The TFEU does not define ‘capital’, although a non-exhaustive list of capital movements can be found in the Annex of Directive 88/361. The following have been held to fall within the provisions on capital movements:

  • Shares. In Commission v Portugal (Case C–367/98) [2002] ECR I–4731, where Portuguese rules precluded investors from other EU Member States from acquiring more than a fixed number of shares in Portuguese companies, the ECJ held that the rules constituted a restriction on capital movements. The ECJ extended this concept in Commission v UK (Case C–98/01) [2003] ECR I–4641, by holding that national rules which restricted share ownership without distinction (i.e. the restriction applied to British nationals as well as foreign investors) were capable of restricting capital movements contrary to EU law.
  • Loans. In Sandoz (Case C–439/97) [1999] ECR I–7041, the ECJ held that provisions of Austrian law which imposed stamp duty, a type of tax, on all loans taken out by Austrian residents, even if the lender was based in another EU Member State, breached EU law on capital movements. The Court ruled that this law would deter Austrian residents from seeking to obtain loans from lenders based in other EU Member States, some of which did not charge stamp duty. Another case in which national provisions which restricted nationals from obtaining loans from foreign providers was Commission v Belgium (Case C–478/98) [2000] ECR I–7587, where the Court stated that ‘measures taken by a Member State which are liable to dissuade its residents from obtaining loans or making investments in other Member States constitute restrictions on the free movement of capital’.
  • Mortgages. In Trummer and Meyer (Case C–222/97) [1999] ECR I–1661, the ECJ held that provisions of Austrian legislation which prohibited the registration of mortgages in the currency of another Member State breached EU law on capital movements.
  • Acquisition of real property, i.e. land. For example, in Konle (Case C–302/97) [1999] ECR I–3099, where Austrian legislation forced all foreign nationals purchasing property in certain tourist-friendly regions of Austria to obtain authorisation first, the ECJ held that a breach of EU rules on capital movements had occurred. The same result was reached in Albore (Case C–423/98) [2000] ECR I–5965, involving Italian legislation which required foreign nationals to obtain authorisation prior to acquiring property in areas of military importance, and in Ospelt (Case C–452/01) [2003] ECR I–9743, involving Austrian legislation requiring prior authorisation to be obtained before purchasing agricultural land. The cases of SEGRO; Horváth (Cases C–52 and 113/16) [2018] 2 CMLR 36 and Grossmania (Case C–177/20), [2022] 3 CMLR 7 involved Hungarian legislation which provided that any rights of ‘usufruct’ (the right to use land, typically for farming) created ‘by a contract between persons who are not close members of the same family’ were to ‘be extinguished by operation of law’. Under this legislation, Günther Horváth, an Austrian national, and the German-owned Grossmania company lost their previously held rights of usufruct over various parcels of agricultural land in Hungary. The ECJ held that the Hungarian legislation was prohibited by Article 63.
  • Inheritances. An inheritance, whether of money, personal property or ‘immovable property’ (land and buildings), is a movement of capital – provided there is some cross-border element. Examples include Eckelkamp & Others (Case C–11/07) [2008] ECR I–6845, involving German and Dutch residents who inherited immovable property in Belgium following their mother’s death; Arens-Sikken (Case C–43/07) [2008] ECR I–6887, involving an Italian resident who inherited a share of immovable property in the Netherlands following the death of her husband; and Block (Case C–67/08) [2009] ECR I–883, involving a German resident who inherited money invested in Spain. Jäger (Case C–256/06) [2008] ECR I–123 was slightly different, in that it involved a French resident inheriting an estate comprising agricultural land and forest in France following the death of his mother. However, the Court found a cross-border element: the mother was living in Germany at the time of her death.
  • Gifts. In Persche (Case C–318/07) [2009] ECR I–359, the ECJ held that provisions of German legislation allowing taxpayers to deduct from their income the cost of gifts to charitable, benevolent or church organisations, but only if the recipient of the gift was a body established in Germany, breached EU law on capital movements. Gifts were a form of capital whether the gift involved money or some item of property. Similarly, in Commission v Austria (Gifts to Research and Teaching Institutions) (Case C–10/10) [2011] ECR I–5389 the Court held that Austrian legislation which allowed taxpayers to deduct the cost of gifts to research and teaching institutions, but only where those institutions were established in Austria, was in breach of Article 63.
  • Dividends. In Santander Asset Management & Others (Cases C–338/11 to C–347/11) [2012] 3 CMLR 12, the Court held that French legislation which imposed tax on dividends paid out by French companies, but only where the shareholders were based outside France (i.e. French-based shareholders did not have to pay the tax) was prohibited by Article 63. In Real Vida Seguros (Case C–449/20) (2021), unreported, the ECJ held that Portuguese legislation which imposed tax on dividends at different rates, depending on whether the shares were listed on the Portuguese stock exchange or on foreign stock exchanges, with the latter subject to a higher tax burden, was contrary to Article 63.

Derogations

Article 65 specifies certain derogations from the free movement of capital provisions. It states, inter alia, that Member States are entitled:

to take all requisite measures to prevent infringements of national law and regulations, in particular in the field of taxation and the prudential supervision of financial institutions, or to lay down procedures for the declaration of capital movements for purposes of administrative or statistical information, or to take measures which are justified on grounds of public policy or public security.

Considering this provision in Association Eglise de Scientologie de Paris (Case C–54/99) [2000] ECR I–1335, the Court stated that these derogations had to be interpreted strictly, could not be used to serve purely economic ends and were subject to the ‘proportionality’ doctrine. The derogation was successfully relied upon in Sandoz (1999) (noted above). The ECJ accepted the argument of the Austrian government that the requirement to pay stamp duty on all loans was designed to prevent ‘taxable persons from evading the requirements of domestic tax legislation through the exercise of freedom of movement of capital’. However, the derogation was unsuccessfully raised in Commission v Belgium (2000), the Court holding that Belgian rules preventing all Belgian residents from obtaining loans abroad was an excessive and disproportionate restriction on the free movement of capital.

In addition, the four-part test set out in Gebhard (1995) (see Chapter 13) has been held to apply to capital movements. The Austrian legislation in Konle (1999) was found to be justifiable (after certain modifications to remove the discriminatory aspects) because it helped the authorities to ensure that large numbers of properties in tourist-friendly areas were not being bought as second or ‘holiday’ homes, whether by foreign nationals or by Austrian nationals looking to escape from the cities and suburbia. This in turn helped to ensure that those areas retained a ‘permanent population and an economic activity independent of the tourist sector’ (in other words, it ensured that those areas did not turn into ‘ghost towns’ when there were no tourists).

The Austrian rules in Ospelt (2003) were also held to be justifiable, on the grounds that they helped to preserve:

agricultural communities, maintaining a distribution of land ownership which allows the development of viable farms and sympathetic management of green spaces and the countryside as well as encouraging a reasonable use of the available land by resisting pressure on land and preventing natural disasters.

However, in both cases the Court found that the requirement of authorisation went beyond what was necessary to achieve its objective. The Court thought that there were other mechanisms available to the national authorities to ensure compliance with planning legislation in tourist-friendly and/or agricultural areas, which would achieve the same objective as authorisation but impose less of an obstacle to capital movements in doing so.

In Commission v Austria (2011), the ECJ rejected two potential justifications for the Austrian legislation in that case. Although the Court accepted that the promotion of research and development was a potential justification, it was not available where it concerned only research and development at the national level. The promotion of national education and training was another potential justification, but the legislation was disproportionate. The onus was on the Austrian government to show that the objective could not be achieved by less restrictive measures, and it had failed to do so.

Economic and Monetary Union

Economic and Monetary Union (EMU) dates back to 1969, when the six heads of state of the EEC agreed a plan, which would have achieved such a union by 1980, although the plan did not work. Nevertheless, by the late 1980s some progress had been made with the establishment of the Exchange Rate Mechanism (ERM), which replaced the normal fixed rate of exchange between different currencies with a floating rate for those states in the ERM. The system was relatively stable and successful until ‘Black Wednesday’ in October 1992, when the UK (and Italy) was forced to withdraw at the cost of billions of pounds to the British Treasury. However, this disaster prompted the Member States to concentrate on taking EMU to the next level: a single currency. First, the various advantages and disadvantages had to be examined.

Advantages and disadvantages of a single (European) currency
  • Advantages: ‘transaction costs’ incurred in the form of commission as one currency is exchanged for another disappear; greater transparency in terms of price comparison between the same or similar goods on sale in different countries; elimination of exchange rate fluctuations should improve confidence for producers and retailers when setting prices.
  • Disadvantages: Member States are often at different points in the economic cycle (i.e. some have growing economies and some have shrinking economies, with differing levels of inflation) but a single currency prevents them from responding individually to these issues, e.g. by unilaterally raising or lowering interest rates.

After weighing up these advantages and disadvantages, most of the EU’s Member States decided to forge ahead. Over the 1990s, these states began to co-ordinate their economic policies and forge closer co-operation between their central banks. A body called the European Monetary Institute (EMI) was established to help. The Treaty on European Union established a set of ‘convergence criteria’ which would be used to determine which states would be eligible to join the single currency. These included:

  • a high degree of price stability, i.e. a rate of inflation close to the three best-performing Member States;
  • sustainability of government deficit at not more than 3% of gross domestic product (GDP) and government debt at not more than 60% of GDP;
  • the observance of the fluctuation margins permitted by the ERM for at least two years.

The criterion concerning government deficit and debt did not just apply at the point of Member States joining the single currency – it continues to apply afterwards. In 1997, the Member States agreed, in the Stability and Growth Pact (SGP), to maintain ‘budget discipline’, i.e. to respect the requirement to keep government deficit at or below 3% of GDP and, if necessary, take immediate corrective action (subject to economic sanctions for failure to do so).

Eleven Member States (Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, the Netherlands, Portugal and Spain) satisfied these convergence criteria, and in January 1999 the European Currency Unit (ECU, subsequently renamed the euro) came into operation in these states for electronic payments. At the same time, the EMI became the European Central Bank (ECB). The ECB, which is based in Frankfurt in Germany, has several tasks including defining and implementing monetary policy and issuing euro banknotes (individual states issue euro coins).

National currencies continued to be used for all other transactions. Greece was deemed to have satisfied the criteria in 2001, and the so-called euro zone grew to 12 members. On New Year’s Day 2002, euro notes and coins came into operation and, two months later, national currencies of the 12 participating states were finally withdrawn. Thus, the Austrian schilling, Belgian and French francs, Dutch guilder, German mark, Greek drachma, Irish punt, Italian lira, Portuguese escudo and Spanish peseta all ceased – overnight – to be legal tender. Slovenia became the 13th Member State to join the ‘euro zone’ in 2007, followed by Cyprus and Malta in 2008, Slovakia in 2009, Estonia in 2011, Latvia in 2014 and Lithuania in 2015. At the time of writing there are therefore 19 Member States in the ‘euro zone’, representing over 340 million people. Eight Member States – Bulgaria, Croatia, Czech Republic, Denmark, Hungary, Poland, Romania and Sweden – remain outside the ‘euro zone’ although they have the option of joining in the future (subject to satisfying the ‘convergence criteria’).

10.2.3 The relationship between ‘services’ and ‘goods’

No definition of ‘goods’ is provided in the TFEU, or in any provisions of secondary legislation. The task has been left to the ECJ. In Commission v Italy (Case 7/68) [1968] ECR 617, the ECJ defined ‘goods’ very widely:

Although this definition is wide, there are some limits. For example, narcotic drugs are not ‘goods’. In B.S. & C.A. (Case C–663/18) [2021] 2 CMLR 5, the Court had to decide whether cannabidiol, a liquid derived from the cannabis plant, qualified as ‘goods’ for the purposes of Article 34. B.S. and C.A. were the directors of a French company, Catlab. In 2014, the company began selling the Kanavape e-cigarette containing cannabidiol in France. Under French legislation, the Public Health Code, the sale of cannabidiol was permitted in France but only if the liquid was derived exclusively from the fibre and seeds of the cannabis plant, otherwise it was classed as a ‘narcotic substance’, the sale of which was illegal in France. The Kanavape was lawfully produced in the Czech Republic; however, the entire plant was used in the production process (i.e. not just the fibre and seeds). In 2018, B.S. and C.A. were prosecuted for selling a ‘narcotic substance’ under the Public Health Code, fined €10,000 each and given suspended prison terms. They appealed and their case was referred to the Court of Justice.

The Court first noted that, unless authorised by the national authorities for medical and scientific purposes, narcotics (such as cannabis) are not ‘goods’, following Josemans (Case C–137/09) [2010] ECR I–13019 (discussed in Chapter 13). Narcotics ‘are prohibited from being released into the economic and commercial channels of the EU’ and ‘persons who market those products cannot rely on the freedoms of movement or the principle of non-discrimination’. Next, the Court acknowledged that a literal interpretation of the United Nations’ Single Convention on Narcotic Drugs (1961), under which ‘cannabis, cannabis resin, extracts and tinctures of cannabis’ are all listed as narcotic drugs, might lead to the conclusion that, insofar as cannabidiol is extracted from cannabis plants, it therefore constituted a ‘cannabis extract’ and, consequently, a ‘narcotic drug’. However, the Court then stated that, because the cannabidiol in the present case did not appear to have ‘any psychotropic effect or any harmful effect on human health on the basis of available scientific data’, it would be ‘contrary to the purpose and general spirit of the Single Convention to include it under the definition of ‘narcotic drugs’. It therefore followed that the cannabidiol in the present case did qualify as ‘goods’ under EU law. (This case will be discussed further in Chapter 14.)

There is a definition of ‘services’ in the TFEU, although it is a far from comprehensive definition. Article 57 states that the word includes ‘activities of an industrial character’, ‘activities of a commercial character’, ‘activities of craftsmen’ and ‘activities of the professions’.

In a number of cases the ECJ has had to decide which of the freedoms of ‘services’ or ‘goods’ applies to the facts. Customs and Excise Commissioners v Schindler (Case C–275/92) [1994] ECR I–1039 is illustrative of the problem facing the ECJ when it is not immediately obvious whether the free movement of goods provisions should apply or whether those on services are more appropriate.

Conversely, in Jägerskiöld v Gustafsson (Case C–97/98) [1999] ECR I–7319, the ECJ held that the provisions on services, and not those on goods, applied in a case concerning disputed fishing permits. Although the permits had a tangible existence, and could be regarded as a product with a monetary value, the ECJ decided that this was ancillary to the service which the permit allowed its holder to carry out.

A similar decision was reached in FA Premier League & Others; Murphy (Cases C–403, 429/08) [2012] 1 CMLR 29. Karen Murphy, a publican in Portsmouth, obtained and used a Greek ‘decoding’ card which enabled her to broadcast, in her pub, Greek satellite TV broadcasts of live English Premier League football matches. As a result she was convicted of breaching UK legislation (s297(1) of the Copyright, Designs & Patents Act 1988), but appealed to the High Court, which referred the case of the ECJ. Given that the case involved the use of an imported product (a decoding card) in order to receive a service (a satellite TV broadcast) from another Member State, the first issue for the Court was whether to apply the Treaty provisions on goods or services. It decided that the provisions on services applied, holding:

where legislation concerns an activity in respect of which the services provided by the economic operators are particularly prominent, whilst the supply of equipment is related thereto in only a purely secondary manner, it is appropriate to examine that activity in the light of the freedom to provide services alone.

(This case is examined in more detail in Chapter 13.)

10.2.4 The relationship between ‘capital’ and ‘goods’

Current legal tender in the form of banknotes and coins is subject to the provisions on the free movement of capital rather than the provisions on the free movement of goods (Bordessa (1995)). By way of contrast, old coins that are no longer legal tender have been held to fall within the provisions on the free movement of goods (R v Thompson and Others (Case 7/78) [1979] ECR 2247, which is examined in Chapter 14).

In Persche (2009), discussed above, a German taxpayer had donated over €18,000 worth of bed linen, towels, zimmer frames and toy cars (all ‘everyday consumer goods’) to a charitable organisation in Portugal. Under German legislation, he was unable to deduct this expense from his taxable income because the organisation was based in another Member State, whereas he would have been able to do so had the organisation been based in Germany. When he challenged this, the ECJ held that Article 63 applied (and had been breached). The Court rejected a suggestion that the rules on the free movement of goods should have been applied instead. The Court noted that the German legislation applied to all gifts, whether of money or consumer goods (and in the case of the latter, regardless of where the goods were purchased).

SUMMARY

  • The internal market is an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured (Article 26 TFEU).
  • There is potential difficulty in distinguishing the free movement of goods from the free movement of services, but it is important to decide which one applies in any given case because different Treaty provisions will then operate. ‘Goods’ are defined as ‘products which can be valued in money, and which are capable, as such, of forming the subject of commercial transactions’ (Commission v Italy), whereas ‘services’ are defined as simply including ‘activities of an industrial character’, ‘activities of a commercial character’, ‘activities of craftsmen’ and ‘activities of the professions’ (Article 57 TFEU).
  • A similar problem arises when distinguishing the free movement of goods from the free movement of capital. In Bordessa, the ECJ decided that current legal tender in the form of banknotes and coins is subject to the provisions on the free movement of capital rather than the provisions on the free movement of goods. In contrast, old coins that are no longer legal tender fall within the provisions on the free movement of goods (R v Thompson and Others).
  • The free movements are not ‘absolute’ freedoms. The Member States are allowed to derogate, using justifications found in the TFEU. All Four Freedoms are subject to derogations on grounds of public policy and public security. Public health can be used to derogate from the freedoms of goods, persons and services.
  • In addition, the ECJ has introduced a policy, common to all of the freedoms, of allowing Member States to justify national legislation which hinders one or more of the freedoms. The legislation must pass four tests (Gebhard): it must (i) be non-discriminatory; (ii) be justified by imperative requirements in the general interest; (iii) be suitable for the attainment of the objective which it pursues; and (iv) not go beyond what is necessary in order to attain its objective (the ‘proportionality’ doctrine).
  • The free movement of capital provisions are found in Articles 63–66 TFEU.
  • ‘Capital’ includes shares, gifts, inheritances, loans, mortgages and the acquisition of real property.
  • Derogations are available to the Member States (Article 65).
  • The idea of Economic and Monetary Union dates back to the late 1960s and came to fruition in the early twenty-first century with the introduction of the single European currency, the euro, in January 2002. The ‘euro zone’ now comprises 19 Member States.

Further reading

Articles

  • Barnard, C, ‘Fitting the Remaining Pieces into the Goods and Persons Jigsaw?’ (2001) 26 EL Rev 35.
  • Daniele, L, ‘Non-Discriminatory Restrictions to the Free Movement of Persons’ (1997) 22 EL Rev 191.
  • Dunnett, D, ‘Some Legal Principles Applicable to the Transition to the Single Currency’ (1996) 33 CML Rev 1133.
  • Flynn, L, ‘Coming of Age: The Free Movement of Capital Case Law 1993–2002’ (2002) 39 CMLL Rev 773.
  • Greaves, R, ‘Advertising Restrictions and the Free Movement of Goods and Services’ (1998) 23 EL Rev 305.
  • Nic Shuibhne, N, ‘Margins of Appreciation: National Values, Fundamental Rights and EC Free Movement Law’ (2009) 34 EL Rev 230.
  • Oliver, P and Roth, WH, ‘The Internal Market and the Four Freedoms’ (2004) 41 CMLL Rev 407.
  • Prechal, S and De Vries, S, ‘Seamless Web of Judicial Protection in the Internal Market?’ (2009) 34 EL Rev 5.
  • Tryfonidou, A, ‘Further Steps on the Road to Convergence among the Market Freedoms’ (2010) 35 EL Rev 36.

11 Citizenship of the Union

DOI: 10.4324/9781003218562-11

11.1 Introduction

The original EC Treaty signed in 1957 made no reference to any form of European ‘citizenship’. At that time, people were regarded as being citizens of their own Member States, and nothing more. That was to change when the Treaty on European Union came into force in November 1993, and several new provisions were inserted into the EC Treaty. These provisions now form Part Two of the TFEU, Articles 18–25. Article 20 provides (in part) as follows:

11.2 Nationality requirements: Article 20 TFEU

Thus, according to Article 20, in order to be a ‘Citizen of the Union’, it is first necessary to hold ‘the nationality of a Member State’. Nationality is conferred purely by way of the national law of the state concerned. In Micheletti (C–369/90) [1992] ECR I–4239, the ECJ stated that ‘it is for each Member State, having due regard to [Union] law, to lay down the conditions for the acquisition and loss of nationality’. However, ‘nationality’ is wide enough to include dual nationals. This is most obviously the case with people holding the nationality of two Member States (for example, Garcia Avello v Belgium (Case C–148/02) [2003] ECR I–11613, involving dual Belgian-Spanish nationality, and McCarthy (Case C–434/09) [2011] ECR I–3375, involving dual British-Irish nationality). However, it is enough if a person holds the nationality of one of the Member States and that of a non-Member State, as Micheletti (1992) illustrates:

Although Micheletti pre-dates the introduction of the citizenship provisions, the case is still good authority for the proposition that holders of dual nationality (only one of the nationalities being in the EU) is sufficient for Article 20 to apply, and hence Article 21 also. This was confirmed in Collins (Case C–138/02) [2004] 3 WLR 1236, where the ECJ held that a dual Irish-American national could rely upon EU law (specifically Article 45 TFEU, the free movement of workers). Other cases to note are:

  • Z.Z. (Case C–300/11) [2013] 3 CMLR 46; [2013] QB 1136 (involving a dual French-Algerian national);
  • Iida (Case C–40/11) [2013] 1 CMLR 47, [2013] 2 WLR 788 (involving a girl with German, Japanese and US nationality);
  • K (Case C–331/16) [2018] 3 CMLR 26; [2019] 1 WLR 1877 (involving a dual Croatian-Bosnian national); and
  • C.G. (Case C–709/20) [2022] 1 CMLR 26; [2021] 1 WLR 5919 (involving a dual Croatian-Dutch national);
  • Tjebbes & Others (Case C–221/17) [2019] 2 CMLR 35; [2019] QB 1476 (involving a Dutch-Canadian national, a Dutch-Swiss national and a Dutch-Iranian national).

All of these cases will be examined in detail later in this chapter.

11.2.1 Loss of nationality

All of the above cases involved the acquisition of EU citizenship. The ECJ has also addressed the loss of EU citizenship on three occasions to date. Rottmann (Case C–135/08) [2010] ECR I–1449 involved a German national who was threatened with having his (acquired) nationality revoked. (He had been born an Austrian national but had later acquired German nationality, giving up his Austrian nationality as he did so. It subsequently emerged that he had not been entirely honest with the German authorities at the time he acquired his new nationality.) As this would also entail him losing his EU citizenship status, the ECJ was asked to provide guidance. The Court held that, in general, Member States were free to decide on matters of nationality, including loss of nationality – a point confirmed in several international conventions. The Court stated that it was ‘legitimate for a Member State to wish to protect the special relationship of solidarity and good faith between it and its nationals and also the reciprocity of rights and duties, which form the bedrock of the bond of nationality’. However, the Court also held that it was incumbent on the national court to assess whether the decision to revoke nationality in any particular case was ‘proportionate’. In deciding this, it would be necessary ‘to take into account the consequences that the decision entails for the person concerned and, if relevant, for the members of his family with regard to the loss of the rights enjoyed by every citizen of the Union’.

Tjebbes & Others (2019) involved a challenge to two provisions of the Dutch Nationality Act. Article 15 stated that an adult Dutch national would lose their nationality if (i) they were a dual national and (ii) had their principal residence for an uninterrupted period of 10 years as an adult outside the Netherlands and its overseas territories. Article 16 stated that a minor child with Dutch nationality would lose that nationality if their parent(s) lost their Dutch nationality. The challenge to Article 15 was brought by three women who were told that they had lost their Dutch nationality after living abroad for more than 10 years: Mrs M. Tjebbes, a Dutch-Canadian national who lived in Canada; Mrs G. Koopman, a Dutch-Swiss national who lived in Switzerland; and Mrs E. Saleh Abady, a Dutch-Iranian national who lived in Iran. The challenge to Article 16 was brought by Miss L. Duboux, Mrs Koopman’s minor daughter, who faced losing her Dutch nationality. The ECJ held that, in principle, Article 20 TFEU did not prevent national law from withdrawing nationality in situations such as those set out in the Dutch Nationality Act. Member States were entitled to

take the view that nationality is the expression of a genuine link between it and its nationals, and therefore to prescribe that the absence, or the loss, of any such genuine link entails the loss of nationality. It is also legitimate for a Member State to wish to protect the unity of nationality within the same family.

However, this was subject to the proportionality principle, which meant that individuals who faced losing their nationality were entitled to ‘an individual examination of the consequences of that loss for the persons concerned from the point of view of EU law’. That was a matter for the national court.

The most recent case is J.Y. (Case C-118/20) [2022] 2 CMLR 23.

11.3 Citizens’ rights of free movement and residence: Article 21(1) TFEU

In Kaur (Case C–192/99) [2001] ECR I–1237 the ECJ confirmed that it was for each state to determine which persons are entitled to nationality of that state and, in turn, whether such nationality actually entitled the holder of it to entry into and residence within the state in question.

This decision must be correct. If Ms Kaur was regarded as a ‘Citizen of the Union’ purely because of her status as a ‘British Overseas Citizen’, then she would have rights to move around the EU and reside in any state of her choosing – except the UK, where British law did not confer on her a right of residence. Such an outcome would be deeply illogical. It must be the case that, before Article 21 can be invoked, the claimant must first have a right of residence in one of the Member States under its national law.

In Lounes (Case C–165/16) [2018] 2 CMLR 9; [2018] QB 1060, the ECJ was required to consider what rights under EU law were available to a national of one Member State who moved to another State (utilising their free movement rights under Article 21) and who subsequently acquired the nationality of the host State.

11.3.1 The ‘direct effect’ question: can Article 21(1) be relied upon in national courts?

In Baumbast (Case C–413/99) [2002] ECR I–7091, the ECJ held that Article 21(1) could be relied upon by EU citizens, to claim a right of residence in other EU Member States. Although Article 21(1) was subject to certain limitations and conditions, this did not deprive it of direct effect. The ECJ held that the competent authorities of the Member State and, where necessary, national courts, must ensure that those limitations and conditions were applied in compliance with the general principles of EU law and, in particular, the principle of proportionality.

11.3.2 Who is entitled to rely upon Article 21(1)?

The case of Baumbast (2002) therefore establishes that Article 21(1) is directly effective, and confers on anyone holding the nationality of a Member State a legally enforceable right of movement to, and residence in, any other EU Member State. That on its own is an important decision, as it apparently obviates the need for persons to establish some form of economic activity (whether as worker, work-seeker, self-employed person, service-provider or service-recipient) in order to claim residence rights under EU law. Over the years, the ECJ has allowed Article 21(1) to be relied upon by a variety of people who, prior to the introduction of citizenship rights, may have struggled to build a case using the more traditional free movement provisions (Articles 45, 49 and 56 – workers, the self-employed and service-providers). The cases to date can be grouped into the following categories:

  • the unemployed – Martínez Sala (Case C–85/96) [1998] ECR I–2691; Collins (2004); De Cuyper (Case C–406/04) [2006] ECR I–6947;
  • students – Grzelczyk (Case C–184/99) [2001] ECR I–6193; D’Hoop (Case C–224/98) [2002] ECR I–6191; Bidar (Case C–209/03) [2005] ECR I–2119; Morgan and Bucher (Cases C–11 & 12/06) [2007] ECR I–9161; Förster (Case C–158/07) [2008] ECR I–8507;
  • children – Zhu and Chen (Case C–200/02) [2004] ECR I–9925; Schwarz (Case C–76/05) [2007] ECR I–6849; Ruiz Zambrano (Case C–34/09) [2011] ECR I–1177; Alokpa & Moudoulou (Case C–86/12) [2017] 1 CMLR 40; Rendón Marín (Case C–165/14) [2017] 1 CMLR 29; [2017] 2 WLR 117;
  • the retired – Pusa (Case C–224/02) [2004] ECR I–5763; Turpeinen (Case C–520/04) [2006] ECR I–10685; Zablocka-Weyhermüller (Case C–221/07) [2008] ECR I–9029; Rüff1er (Case C–544/07) [2009] ECR I–3389; Brey (Case C–140/12) [2014] 1 CMLR 37; [2014] 1 WLR 1080;
  • those incapable of working for health reasons – Tas-Hagen (Case C–192/05) [2006] ECR I–10451; Nerkowska (Case C–499/06) [2008] ECR I–3993;
  • amateur sportsmen and women – TopFit & Biffi (Case C–22/18) [2020] 1 CMLR 3.

The crucial threshold that has to be crossed in all citizenship cases is that ‘beneficiaries of the right of residence must not become an unreasonable burden on the public finances of the host Member State’ (Grzelczyk (2001)). That criterion itself begs the (as yet unanswered) question – what exactly is an ‘unreasonable burden’? In several of the Article 21 cases, EU citizens have been held to be entitled to claim social security or other financial benefits such as:

  • a child-raising allowance – Martínez Sala (1998);
  • a ‘minimum subsistence allowance’ – Grzelczyk (2001);
  • a ‘tideover allowance’ – D’Hoop (2002);
  • a student loan – Bidar (2005);
  • a student grant – Förster (2008);
  • a ‘compensatory supplement’ – Brey (2014).

So, it is clear that, just because the EU citizen is in need of financial help from the state, this does not make him or her an ‘unreasonable burden’.

11.3.3 The right of citizens to invoke Article 18 TFEU to challenge discrimination based on nationality

A right of residence alone is potentially of little assistance unless it is accompanied by a right not to be discriminated against. Article 21(1) itself says nothing about such a right, but the ECJ has allowed citizens to invoke Article 18 TFEU in order to challenge discrimination based on nationality. Article 18 states:

The earliest case was Martínez Sala (1998).

The Court stated:

Martínez Sala was applied in Grzelczyk (2001), a case involving a French national in his final year as a student at a university in Belgium. He submitted a claim for financial assistance from the Belgian social service, known as the minimex, to allow him to concentrate on his dissertation rather than having to take part-time jobs. However, this was turned down, on the basis that he was neither a Belgian national nor a migrant worker, as required by Belgian legislation. The ECJ, however, held that, as a French national lawfully resident in Belgium, he was protected by Article 21(1) and allowed to invoke Article 18 to claim the minimex. The Court stated:

Other cases involving students invoking the combined effect of Article 21(1) (residence) and Article 18 (non-discrimination) include D’Hoop (2002), considered below, and Bidar (2005).

11.3.4 The right of citizens to invoke Article 21(1) to challenge national legislation which places them ‘at a disadvantage’

In a number of cases the ECJ has upheld claims brought by EU citizens challenging their home state’s legislation based purely on Article 21(1), on the basis that the legislation had placed them ‘at a disadvantage’ because they had ‘exercised their freedom to move and to reside in another Member State’. Examples include Pusa (2004), Turpeinen (2006), De Cuyper (2006), Tas-Hagen (2006) and Morgan and Bucher (2007). There is no requirement in such cases to establish that the legislation in question was either directly or even indirectly discriminatory (although if it was, then Article 18 could be invoked as well).

One of the cases on students, Morgan and Bucher (2007), illustrates this point. The case involved two German students at educational establishments in other Member States. Rhiannon Morgan was studying at the University of the West of England in Bristol, while Iris Bucher was a student at the Hogeschool Zuyd in Heerlen, in the Netherlands. Both women applied for a training grant from the authorities in Germany, but in each case it was turned down on the basis that, although funding was available for study abroad, this was only the case where students were continuing studies that had been commenced in Germany. Both women appealed, and the ECJ agreed that in each case the women, being German nationals lawfully resident in the UK and the Netherlands, respectively, could invoke Article 21(1). The Court ruled that the restriction of funding for overseas students was simply a breach of Article 21(1) itself. The Court stated:

This development can also be seen in Pusa (2004) and Turpeinen (2006), both involving retired Finnish nationals who went to live in Spain (presumably to take advantage of the more clement weather in Iberia compared to Scandinavia). The situation where retired persons seek to invoke citizenship rights can only increase in significance in the future, given improving health and living conditions, so these cases are clearly very significant. Both Heikki Pusa and Pirkko Turpeinen had encountered problems after having moved to Spain. Essentially the disputes involved their Finnish retirement pensions being subjected to greater tax liability, as a result of their move to Spain, than would have been the case had they remained in Finland. In each case the ECJ held that Article 21(1) was available, in principle at least, to challenge any disadvantage incurred as a result of exercising their freedom of movement rights. In Pusa, the Court stated:

The judgment in Turpeinen was worded almost identically. Another, similar, category of person seeking to invoke citizenship rights under Article 21(1) is those who have been forced to give up working because of health reasons. The leading case to date is Tas-Hagen (2006).

Therefore, Tas-Hagen decides that a restriction on the availability of civilian war benefit can be challenged using Article 21(1). Now, civilian war benefit is not something that falls within the scope of EU law generally (compare the case of Baldinger (Case C–386/02) [2004] ECR I–8411, discussed in Chapter 12, where war compensation was held not to fall within the scope of Regulation 1612/68). However, that appears to have no importance where an Article 21(1) claim is involved. Yuri Borg-mann-Prebil (‘The Rule of Reason in European Citizenship’ (2008) 14 ELJ 328) argues this point as follows (emphasis added):

11.3.5 Citizenship rights can be invoked in the citizen’s ‘home’ state

In D’Hoop (Case C–224/98) [2002] ECR I–6191 the ECJ decided that Article 21(1) could be relied upon in the citizen’s home state, provided that she had gone to another EU Member State in order to exercise her free movement rights before returning.

11.3.6 Children’s rights to invoke Article 21(1)

In Zhu and Chen (Case C–200/02) [2004] 3 WLR 1453, the question arose as to (1) whether an infant child could rely upon her status as a ‘Citizen of the Union’ in order to claim a right of residence in another EU Member State and, if so, (2) whether her mother, as ‘primary carer’, was entitled to residence with her. The ECJ answered both questions in the affirmative.

The Court stated:

Another case in which children were held to be entitled to invoke Article 21(1) is Schwarz (Case C–76/05) [2007] ECR I–6849. The Schwarz family lived in Germany. In 1998, Herbert Schwarz and his wife Marga decided to send two of their three teenage children to the Cademuir International School in Scotland, which specialised in exceptionally gifted children. Under German law, however, this meant that they could not claim tax relief on school fees (around 10,000 DM per year) which would have been available had the children been schooled privately in Germany. They challenged this decision, invoking Article 56 (the freedom to receive services) or, alternatively, Article 21(1). The ECJ decided that the case would fall within Article 56 if the Cademuir School was ‘essentially financed by private funds’, this being a question of fact for the national court. If not, then Article 21(1) would apply. The Schwarz children had exercised their free movement rights, notwithstanding their youth (following Zhu and Chen). In either case, there was a breach, because there was no justification for the discrimination between schooling in Germany and schooling abroad for the purposes of conferring tax relief on school fees. On the ability of children to invoke Article 21(1), the Court stated:

N.B. The German deutschmark ceased to be legal tender when the single European currency was introduced in 2002. Other ECJ cases involving the exercise by children of their citizenship rights under EU law are Bajratari (Case C–93/18) [2020] 1 CMLR 29; [2020] 1 WLR 2327, which will be discussed in section 11.4.4, and Rendón Marín (Case C–165/14) [2017] 1 CMLR 29; [2017] 2 WLR 117, which will be discussed in section 11.6.8.

11.3.7 Justification for restrictions on citizens’ rights

In many of the cases discussed above, the ECJ ruled that EU citizens could invoke Article 21(1) (either on its own or in conjunction with Article 18) in order to challenge national legislation which was either discriminatory or places the citizen ‘at a disadvantage’ as a result of he or she exercising their free movement rights. However, in most of those cases, that was not the end of the story. The Court is prepared to listen to arguments that national legislation imposing restrictions on a citizen’s free movement rights constituting a prima facie breach of Article 21(1) can be justified by referring to ‘objective considerations of public interest’. For example, in Tas-Hagen (2006), discussed above, the Court stated:

It is essential to note the four elements of this extract from the Tas-Hagen judgment. In order for a restriction on a citizen’s rights to be justified it must be:

  • ‘Based on objective considerations of public interest’ – the cases of D’Hoop (2002), Bidar (2005) and Tas-Hagen itself, discussed below, provide examples of such ‘objective considerations’.
  • ‘Independent of the nationality of the persons concerned’ – the national legislation must either be non-discriminatory or, at the most, ‘indirectly’ discriminatory. This typically means legislation that, at first sight, appears not to discriminate but in practice has that effect. The most common example is a residency requirement. The cases of Bidar and Tas-Hagen both illustrate this situation. See also section 12.4.2 for more examples of ‘indirect’ discrimination.
  • ‘Appropriate’ – the restriction must be capable of achieving the desired objective.
  • ‘Proportionate’ – the restriction must be absolutely necessary. In other words, it should impose the minimum restriction needed in order to achieve the desired objective, and not go any further than that.

D’Hoop (2002), discussed above, illustrates these points. First, the Court accepted that it was ‘legitimate for the national legislature to wish to ensure that there is a real link between the applicant for that allowance and the geographic employment market concerned’. In other words, the requirement that claimants had undertaken their schooling in Belgium was potentially justifiable in order to establish a ‘real link’. Second, the legislation did not expressly refer to nationality. However, the Belgian legislation failed the appropriateness test. The schooling requirement was not capable of establishing the ‘real link’, being ‘too general and exclusive in nature’.

Bidar (2005) further illustrates this point. In that case the ECJ accepted, in principle, that the UK legislation imposing a residency requirement on applicants for a student loan was prima facie justifiable. The Court held that it was ‘legitimate’ for the UK to seek to ensure:

that an applicant for assistance has demonstrated a certain degree of integration into the society of that State [which] may be regarded as established by a finding that the student in question had resided in the host Member State for a certain length of time.

However, the UK legislation also failed the appropriateness test. The Court stated:

Tas-Hagen (2006) provides another example. The Dutch government argued that the requirement in Dutch law that claimants for civilian war benefit had to be resident in the Netherlands when the claim was made reflected the ‘legislature’s wish to limit the obligation of solidarity with civilian war victims to those who had links with the population of the Netherlands during and after the war’. Although the ECJ was prepared to accept ‘the obligation of solidarity’ as an ‘objective consideration of public interest’, the ECJ held that the residency criterion was not an appropriate means of establishing ‘solidarity’. The Court stated:

a residence criterion … based solely on the date on which the application for the benefit is submitted, is not a satisfactory indicator of the degree of attachment of the applicant to the society which is thereby demonstrating its solidarity with him.

One case where all four elements were satisfied, including both the appropriateness and proportionality tests, is Förster (2008).

In Patmalniece [2011] UKSC 11, [2011] 1 WLR 783, the UK Supreme Court ruled that a provision in the State Pension Credit Act 2002 which required claimants to be ‘habitually resident’ in Great Britain constituted indirect discrimination. However, the Court accepted the government’s argument that it was justifiable, because of the need ‘to prevent exploitation of welfare benefits by people who come to this country simply to live off benefits without working here’ (according to Lord Hope). More recently, in Commission v UK (Social Security Benefits) (Case C–308/14) [2016] 3 CMLR 41; [2016] 1 WLR 5049, the ECJ held that UK legislation (the Social Security Contributions and Benefits Act 1992 and the Tax Credits Act 2002), which required that people claiming child benefit or child tax credit, respectively, had to be ‘ordinarily resident’ in the UK was indirect discrimination against EU citizens. However, such discrimination was justifiable based on ‘the need to protect the finances of the host Member State’.

11.4 Citizens’ rights of exit, entry and residence: Directive 2004/38

The rights of a citizen to leave the territory of his or her ‘home’ state, to enter the territory of the ‘host’ state, to bring their family with them and to set up home there, and even to retire there after their working life is over, are all covered by Directive 2004/38. This repealed pre-existing legislation, Directive 68/360, although some of the case law on that directive is still of relevance today. Indeed, some of that case law has been incorporated into the new directive. This section will provide an overview of the main provisions of the new directive.

11.4.1 Rights of exit: Article 4

Article 4 deals with the right of a citizen to leave their ‘home’ state in order to work in another state (the ‘host’ state). Article 4(1)–(3) provides as follows:

(Note: the meaning and scope of a citizen’s ‘family members’ will be discussed in the next section of this chapter [section 11.5].) In A (Case C–35/20) (2021), unreported, the ECJ held that Article 4(1) did not confer absolute free movement. Member States were entitled to require their own Citizens to be in possession of their ID documents when leaving the jurisdiction ‘by whatever means of transport and by whatever route’. In the case, A, a Finnish national, had gone on a day trip on a pleasure boat from Helsinki. During the trip, the boat crossed into Estonian waters before returning to Helsinki. On his return, A was asked for his passport. He had not taken his passport with him and was prosecuted under Finnish law for failing to carry his ID card or passport when leaving Finland. A challenged that but the Court upheld the Finnish law, provided that whatever punishment was imposed was proportionate.

11.4.2 Rights of entry: Article 5

Article 5(1) provides that Member States must grant to all Union citizens the right to enter their territory with a valid ID card or passport, and must also grant family members who are non-EU nationals the right to enter their territory with a valid passport. Host Member States cannot require an entry visa from Union citizens (Article 5(1)). Member States may require a visa for non-EU national family members, in certain circumstances, but ‘possession of the valid residence card referred to in Article 10 shall exempt such family members from the visa requirement’ (Article 5(2)). Article 5(2) was considered in McCarthy & Others (Case C–202/13) [2015] 2 CMLR 13; [2015] QB 651.

Article 5(4) provides:

Finally, Article 5(5) provides that the host Member State may require the person concerned to report his or her presence within its territory within a ‘reasonable and non-discriminatory period of time’. Failure to comply with this requirement may make the person concerned ‘liable to proportionate and non-discriminatory sanctions’.

11.4.3 Rights of residence for up to three months: Article 6

Article 6(1) provides that Union citizens shall have the right of residence in the ‘host’ state for a period of up to three months without any ‘conditions’ or ‘formalities’ other than the requirement to hold a valid ID card or passport. Article 6(2) applies the provisions of Article 6(1) to non-EU family members who are ‘accompanying or joining’ the Union citizen, provided that they are in possession of a valid passport.

Article 6(1) was considered in F.S. (Case C–719/19) [2022] 1 CMLR 7, involving a Polish national who had been ordered to leave the Netherlands as he did not satisfy any of the residency criteria in Article 7(1) (set out in the next section) and had also been arrested on several occasions for shoplifting and pickpocketing. He left the Netherlands voluntarily and went to Germany, but a few weeks later he was again arrested by Dutch police for shoplifting. F.S. claimed that by leaving the Netherlands and returning, he had acquired a new right of residence under Article 6(1). The ECJ held that, once a Citizen had left a Member State of which they were not a national, having been required to do so by the host State authorities, they were not entitled to immediately return and claim a new residence right under Article 6(1). To do so would distort the objective of the Directive, which was to allow for temporary residence under Article 6(1), indefinite residence under Article 7(1) and permanent residence under Article 16(1). Instead, in order to claim a new right of residence under Article 6(1), the Citizen:

must not only physically leave that territory, but also have genuinely and effectively terminated his or her residence on that territory, with the result that, upon his or her return to the territory of the host Member State, his or her residence cannot be regarded as constituting in fact a continuation of his or her preceding residence on that territory. [emphasis added]

In deciding whether the Citizen had ‘genuinely and effectively’ terminated their residence in the host State, account had to be taken of the extent to which they were ‘integrated’ in the host State, the length of their residence in that State, and their ‘family and economic situation’. Other factors to be taken into account included a ‘request for removal from a population register, the termination of a lease contract or a contract for the provision of public services, such as water or electricity, moving house or flat, de-registration from a job placement service or the termination of other relationships which presuppose some integration’ of the Citizen into the host State. The Citizen’s conduct after leaving the host State was also relevant, i.e. ‘evidence suggesting that he or she moved the centre of his or her personal, occupational or family interests to another State during that period’. However, there was no minimum period (such as three months, as had been proposed by the Dutch government) during which the Citizen would be legally barred from returning to the host State. Rather,

the longer the person concerned is absent from the host State, the more that absence attests to the genuine and effective nature of the end of his or her residence. By contrast, merely a very brief absence, of a few days or even a few hours, tends rather to establish that the residence relied on by the Citizen, under Article 6, upon his or her return to the host Member State, is in fact part of the same period of residence in that territory.

11.4.4 Rights of residence for more than three months: Article 7

Article 7(1) provides that Union citizens have a right of residence for more than three months in the ‘host’ state if they:

  1. are workers or self-employed;
  2. have ‘sufficient resources’ for themselves and their family members not to become a ‘burden’ on the social assistance system of the ‘host’ state during their period of residence and have ‘comprehensive sickness insurance cover’;
  3. are students, have ‘comprehensive sickness insurance cover’ and ‘assure the relevant national authority’ that they have ‘sufficient resources’.

Article 7(1)(d) adds that the right of residence is also conferred on ‘family members accompanying or joining a Union citizen who satisfies the conditions referred to in points (a), (b) or (c)’. Article 7(2) confirms that this right extends to non-EU family members.

Workers or the self-employed: Article 7(1)(a)

The meaning of the word ‘worker’ and the additional rights enjoyed by workers and their family members will be examined in more detail in Chapter 12; the rights enjoyed by those in self-employment will be examined in more detail in Chapter 13.

Those with ‘sufficient resources’ not to become a ‘burden’: Article 7(1)(b)

The directive does not define the phrase ‘have sufficient resources’ or the word ‘burden’. Inevitably, this has generated case law, which has established a number of points. First, there is no requirement that the ‘resources’ actually belong to the Citizen: the question is whether they are ‘available’ to him or her. In Alokpa & Moudoulou (Case C–86/12) [2017] 1 CMLR 40, the ECJ stated that ‘the expression “have” sufficient resources in Article 7(1)(b) must be interpreted as meaning that it suffices that such resources are available to the Union citizens…. That provision lays down no requirement whatsoever as to their origin’ (emphasis added). This meant that two French children, who had moved to Luxembourg with their mother (a non-EU national) could claim a right of residence in that Member State provided that funds were ‘available’ to them, either personally or via their mother. Similarly, in Singh & Others (Case C–218/14) [2016] 1 CMLR 12; [2016] QB 208, the ECJ held that the ‘fact that some part of the resources available to the Union citizen derives from resources obtained by the spouse from his activity in the host Member State does not preclude the condition concerning the “sufficiency” of resources in Article 7(1)(b) from being regarded as satisfied’. That case involved three couples living in Ireland: a Latvian woman and her Indian husband; a German woman and her husband, a national of Cameroon; and a Lithuanian woman and her Egyptian husband. It was held to be irrelevant whether or not the three women, all EU citizens, had resources personally. Provided that their respective husbands (all non-EU nationals) had resources which were ‘available’ to their wives, then the women could invoke Article 7(1)(b).

In Bajratari (Case C–93/18) [2020] 1 CMLR 29; [2020] 1 WLR 2327, the facts were not dissimilar to Alokpa & Moudoulou. Ermira Bajratari, an Albanian national, lived in the UK with her husband (also Albanian) and their three children, two of whom had Irish nationality, having been born in Northern Ireland. Their only source of income was the husband’s earnings. When Ermira applied to the Home Office for recognition of her Irish children’s right of residence this was rejected, on the basis that they personally did not have ‘sufficient resources’. However, the ECJ ruled in Ermira’s favour. The Court held that ‘the fact that the resources available to a Citizen minor derive from income obtained by a third-State national parent’s employment in the host State does not prevent the condition of “sufficient resources” from being regarded as satisfied’. This was confirmed in R.H. (Case C–836/18) [2020] 3 CMLR 13; [2020] 4 WLR 59, involving a Spanish national and her Moroccan husband who were financially dependent on the former’s father. The Court said:

A second principle which has emerged from the case law is that there is no fixed sum of money which has to be available to the citizen in order to qualify as ‘sufficient’ resources. In Brey (Case C–140/12) [2014] 1 CMLR 37; [2014] 1 WLR 1080, the ECJ held that, ‘although Member States may indicate a certain sum as a reference amount, they may not impose a minimum income level below which it will be presumed that the person concerned does not have sufficient resources, irrespective of a specific examination of the situation of each person concerned’. Third, the reference to a ‘burden’ on the social assistance system meant an ‘unreasonable burden’. It followed that the ‘mere fact’ that a citizen had applied for benefits in the host state was ‘not sufficient to show that he constitutes an unreasonable burden on the social assistance system of the host Member State’ (Brey). This meant that Peter Brey, a retired German national who had moved to Austria, was not necessarily a ‘burden’ on the Austrian social assistance system merely because he had applied for a ‘compensatory supplement’ – a form of social security benefit in Austria – in order to top up his income (which comprised his pension and a care allowance for his wife, amounting to just over €1100 per month). However, the fact that a non-economically active citizen, such as Peter Brey, may be eligible to receive a state benefit could be ‘an indication’ that he did not have ‘sufficient resources’ to avoid becoming an ‘unreasonable’ burden. The competent national authorities could not, however, draw such a conclusion without first carrying out an overall assessment of the ‘specific burden’ which granting that benefit would place on the national social assistance system as a whole, by reference to the ‘personal circumstances’ of the citizen concerned.

‘Comprehensive sickness insurance cover’

In A (Sickness Insurance) (Case C–535/19) [2022] 1 CMLR 25, the ECJ held that to allow an economically inactive Citizen to claim free access to the host State’s health service (and with it, rights of residence under Article 7(1)(b)) would run counter to the requirement that Citizens had their own ‘comprehensive sickness insurance’, although the Court added that to bar such Citizens from any access to the service would be discriminatory on nationality grounds. Rather, the Citizen should be allowed to access the service ‘subject to conditions intended to ensure that that Citizen does not become an unreasonable burden on the public finances of that Member State’. Such conditions may include ‘payment, by that citizen, of a contribution to that Member State’s public sickness insurance system’ provided ‘that the principle of proportionality is observed and, therefore, that it is not excessively difficult for that Citizen to comply with such conditions’.

In V.I. (Case C–247/20) [2022] 3 CMLR 17; [2022] 1 WLR 2902, the ECJ confirmed that ‘although the host Member State may make affiliation to its public sickness insurance system … subject to conditions intended to ensure that that Citizen does not become an unreasonable burden on the public finances of that Member State’, it followed that once a Citizen had actually become ‘affiliated to … a public sickness insurance system in the host Member State … he or she has “comprehensive sickness insurance” within the meaning of Article 7(1)(b)’. This contradicted and effectively overruled earlier decisions of the Court of Appeal in which it had been held that access to the UK’s NHS did not satisfy Article 7(1)(b) (Ahmad [2014] EWCA Civ 988; Pelltari v Southwark LBC [2015] EWCA Civ 300). In the V.I. case, the Court also held that, once a right of permanent residence had been acquired under Article 16(1), it was no longer a requirement for the person concerned to have ‘comprehensive sickness insurance’. Article 16(1) is discussed in section 11.4.9.

Retention of residence rights: Article 7(3)

Article 7(3) provides that the right of residence conferred by Article 7(1)(a) is not lost by a worker if he or she:

  • is temporarily unable to work as the result of an ‘illness or accident’;
  • is in ‘duly recorded involuntary unemployment’ after having been employed for more than one year;
  • ‘having become involuntarily unemployed during the first twelve months’ – here, ‘worker’ status is retained for ‘no less than six months’;
  • embarks on ‘vocational training’ – unless he or she is involuntarily unemployed, the retention of the status of worker shall require the training to be related to the previous employment.

Some of the provisions of Article 7(3) are ambiguous, but case law is helping to clarify its scope. Most notably, in St Prix (Case C–507/12) [2015] 1 CMLR 5, the ECJ held that pregnancy did not fall within the scope of the phrase ‘illness or accident’, and thus a woman who was forced to temporarily leave her job in the late stages of pregnancy was not protected by Article 7(3)(a). The Court stated that ‘pregnancy must be clearly distinguished from illness, in that pregnancy is not in any way comparable with a pathological condition’. (However, the Court went on to hold that a pregnant woman could retain ‘worker’ status under Article 45 TFEU instead, subject to certain conditions being met – see Chapter 12.)

In Samin [2016] UKSC 1, [2016] 1 WLR 481, the Supreme Court of the UK held that Wadi Samin, an Austrian national living in the UK, could not invoke Article 7(3) because he had been declared permanently incapacitated from working on health grounds, whereas the Directive only protected those who were ‘temporarily’ unable to work. In Gusa (Case C–442/16) [2018] 2 CMLR 23, the ECJ held that the phrase ‘involuntary unemployment’ covered both employed and self-employed people whose work had ceased ‘for reasons beyond their control’. Finally, in Tarola (Case C–483/17) [2019] 3 CMLR 23, the ECJ accepted that the words ‘during the first twelve months’ could either mean the first 12 months of residence or the first 12 months of work. However, because EU law should not be interpreted restrictively, the Court adopted the latter interpretation.

11.4.5 Registration: Articles 8–11

Article 8(1) provides that, for periods of residence longer than three months, the ‘host’ state may require citizens to register with the relevant authorities. Failure to do so ‘may render the person concerned liable to proportionate and non-discriminatory sanctions’ (Article 8(2)). This point was originally established by the ECJ in a case on Directive 68/360, Royer (Case 48/75) [1976] ECR 497, in which the Court ruled that it would be disproportionate to expel or imprison a migrant worker for failing to register.

Once registered, the host state must then immediately issue to the citizen a ‘registration certificate’. In order to register, Article 8(3) provides that the citizen needs to produce certain documents, depending on the capacity in which he or she is seeking to register. In all cases, a valid ID card or passport is required, plus:

  • confirmation of engagement from the employer or a certificate of employment, or proof that they are self-employed persons (for those under Article 7(1)(a)); or
  • proof that they satisfy the relevant conditions laid down in Article 7(1)(b); or
  • proof of enrolment at an accredited establishment, proof of comprehensive sickness insurance cover and the declaration about ‘sufficient resources’ (for those under Article 7(1)(c)).

Article 8(5) deals with the right of family members (who are also EU nationals) to claim registration certificates. As with the citizen, a valid ID card or passport is required, plus certain additional documents proving the family member’s relationship with the citizen (e.g. as spouse, partner, dependant). Non-EU nationals cannot claim such certificates. Instead, Article 9(1) states that non-EU family members are to be issued with a ‘registration card’. As above, failure to comply with the requirement to apply for a registration card may make the person concerned liable to ‘proportionate and non-discriminatory sanctions’ (Article 9(3)). Article 10 sets out the documents to be provided by the non-EU family member (these are essentially the same as those set out in Article 8(5)). Registration cards ‘shall be valid for five years from the date of issue or for the envisaged period of residence of the Union citizen, if this period is less than five years’ (Article 11(1)).

11.4.6 Effect of death or departure of the citizen: Article 12

If the citizen dies, or departs the ‘host’ state, then his or her family members (who are also EU nationals) do not lose their right of residence in that state (Article 12(1)). The situation for non-EU family members is slightly more complicated. Article 12(2) provides that the death of the citizen shall not entail loss of the right of residence of his or her non-EU family members, provided that they have been residing in the host Member State as family members for at least one year before the citizen’s death. Article 12(3) adds that neither the death of the citizen, nor their departure from the host Member State, shall entail loss of the right of residence of either his or her children, or of the parent who has actual custody of the children, irrespective of nationality, if the children reside in the host Member State and are enrolled at an educational establishment, for the purpose of studying there, until the completion of their studies. Article 12(3) gives effect to the ECJ’s decision in Baumbast (2002), which was discussed above.

11.4.7 Effect of divorce, annulment of marriage or termination of registered partnership: Article 13

Article 13 deals with the entitlement of former husbands, wives and partners to remain in the host state after divorce, marriage annulment or the termination of a registered partnership. Article 13 is discussed in section 11.5.

11.4.8 Retention of residence rights: Article 14

Article 14 deals with retention of residence rights. It states:

Article 14(4)(b) is designed to implement the ECJ decisions in Antonissen (Case C–292/89) [1991] ECR I–745 and Collins (2004), which are discussed in more detail in Chapter 12.

11.4.9 Right of permanent residence: Articles 16–21

Articles 16–18 of Directive 2004/38 deal with the acquisition of a permanent right of residence. In Lassal (Case C–162/09) [2010] ECR I–9217, the Court emphasised the importance of this right, describing it as ‘a key element in promoting social cohesion’ which was provided ‘in order to strengthen the feeling of Union citizenship’. Article 16 sets out the conditions under which citizens and their family members can acquire a right of permanent residence in the ‘host’ state. According to Article 16(1), this right is conferred on all citizens who ‘have resided legally for a continuous period of five years’ in the host state. In Ziolkowski & Szeja (Cases C–424, 425/10) [2011] ECR I–14035, the Court held that ‘resided legally’ meant in accordance with the provisions of Article 7(1); in other words, it was not permissible simply to reside in a Member State for five years and then claim a right of permanent residence.

It logically followed that time spent in prison does not count as ‘legal’ residence for the purposes of Article 16(1). This was confirmed by the ECJ in Onuekwere (Case C–378/12) [2014] 2 CMLR 46; [2014] 1 WLR 2420. The Court also held that time spent in prison interrupted the ‘continuity’ of legal residence and prevented the aggregation of periods before and after imprisonment from counting towards a right of permanent residence. The Court stated:

In Lassal, the ECJ also held that Article 16(1) may be used to claim a right of permanent residence, even though the five years in question was completed prior to the Directive’s implementation date (30 April 2006). This ruling was followed in Dias (Case C–325/09) [2011] 3 CMLR 40 and Ziolkowski & Szeja (2011), in which the Court also held that time spent in a Member State by persons from a non-EU country which subsequently became a Member State counted as residency for the purposes of Article 16(1) – provided, of course, that the Article 7(1) criteria were met. This meant that Tomasz Ziolkowski and Barbara Szeja, both Polish nationals, who had been living in Germany since the late 1980s could invoke Article 16(1) to claim a right of permanent residence in Germany even though Poland did not accede to the EU until May 2004.

Article 16(2) provides for non-EU family members. They may also acquire a right of permanent residence provided they have ‘legally resided with the Citizen in the host Member State for a continuous period of five years’. The requirement proved problematic for the family member in Ojo [2015] EWCA Civ 1301. Miss Ojo, a Nigerian national, entered the UK with her mother (who had acquired Austrian nationality) in July 2007. The two women lived together in the south of England until January 2009 when Miss Ojo moved to Manchester to work. Sixteen months later, Miss Ojo moved back south and returned to live with her mother. However, when Miss Ojo later claimed a right of permanent residence under Article 16(2), the Court of Appeal rejected her claim – she had not resided with her mother ‘for a continuous period of five years’.

Thus both Article 16(1) and (2) require ‘continuous’ residence. This point is amplified in Article 16(3):

Article 16(4) states that, once acquired, the right of permanent residence is lost only through absence from the ‘host’ state for a period exceeding two consecutive years. In Dias (2011), the ECJ noted that a right of permanent residence (once acquired) could be lost, according to Article 16(4), ‘only’ by spending at least two consecutive years out of the country. However, the Court then stated that the ‘integration link’ implied by Article 16(1) could also be broken by someone who had legally resided in the host state for at least five years if that was followed by a period of non-legal residence (such as time spent in voluntary unemployment) of at least two years.

Article 17 deals with certain exemptions from the normal five-year qualification period. Article 17(1) states that the right of permanent residence in the ‘host’ state may be acquired before completion of the continuous five-year residence period laid down in Article 16 by workers or self-employed persons who:

  1. At the time they stop working, have reached the age laid down by the law of that Member State for entitlement to an old age pension or workers who cease paid employment to take early retirement, provided that they have been working in that Member State for at least the preceding 12 months and have resided there continuously for more than three years.
  2. Have resided continuously in the host Member State for more than two years and stop working there as a result of permanent incapacity to work. If such incapacity is the result of an accident at work or an occupational disease entitling the person concerned to a benefit payable in full or in part by an institution in the host Member State, no condition shall be imposed as to length of residence.
  3. After three years of continuous employment and residence in the host Member State, work in an employed or self-employed capacity in another Member State, while retaining their place of residence in the host Member State, to which they return, as a rule, each day or at least once a week.

In Gubeladze [2019] UKSC 31, [2019] 3 WLR 71, the UK Supreme Court held that Article 17(1) only required factual residence, not ‘legal’ residence (unlike Article 16(1)). This meant that Tamara Gubeladze, a Latvian national, who had lived in the UK since 2008 (i.e. well over three years) and had reached the UK’s state retirement age, was able to rely on Article 17(1), without having to demonstrate that her residence was necessarily ‘legal’. In A.T. (Case C–32/19) [2020] 2 CMLR 30; [2020] 1 WLR 2370, the ECJ held that the requirement for the Citizen to have ‘been working in that Member State for at least the preceding 12 months’ applied to both categories of retirees in Article 17(1)(a), i.e. those that ‘have reached the age laid down by the law of that Member State for entitlement to an old age pension’ and those ‘who cease paid employment to take early retirement’.

The remainder of Article 17 deals with family members. Article 17(2) states that the conditions as to length of residence and employment laid down in Article 17(1) (a), and the condition as to length of residence laid down in Article 17(1)(b), shall not apply if the worker’s or the self-employed person’s spouse or partner is a national of the ‘host’ state or lost such nationality on marriage to that worker or self-employed person. Otherwise, Article 17(3) or (4) apply. These provisions are as follows:

Article 17(4)(a) is slightly ambiguous – does the two-year residency have to have occurred at any time in the citizen’s past, or does it have to immediately precede their death? The ECJ considered the meaning of the phrase ‘resided continuously’ in Givane (Case C–257/00) [2003] ECR I–345. The Court stated that the provision was ‘intended to establish a significant connection between, on the one hand, [a] Member State and, on the other hand, [a citizen] and his family, and to ensure a certain level of their integration in the society of that State’. It therefore ruled that the residence period had to occur immediately prior to the citizen’s death.

Article 18 states that the non-EU family members of a citizen may acquire a right of permanent residence in the ‘host’ state after residing there legally for five consecutive years. Article 19 provides that Member States are to issue to citizens who comply with the above provisions a ‘document certifying permanent residence’. Similarly, non-EU family members who comply with the above provisions are to be issued with ‘a permanent residence card’ (Article 20(1)). Article 21 states: ‘Continuity of residence is broken by any expulsion decision duly enforced against the person concerned’ (see section 11.6 for discussion of the situations in which Member States can expel other states’ nationals).

11.4.10 Territorial restrictions: Article 22

Article 22 states:

11.4.11 Equal treatment: Article 24

Article 24(1) provides for a general principle of ‘equal treatment’ for all citizens (plus their family members) with the nationals of the ‘host’ state ‘within the scope of the Treaty’. This provision has generated some important case law, primarily dealing with the question whether economically inactive EU citizens can invoke Article 24(1) in order to claim social security benefits. In the first case, Dano (Case C–333/13) [2015] 1 CMLR 48; [2015] 1 WLR 2519, the ECJ held that the answer was ‘no’: only citizens with a right of residence under Article 7(1) – discussed above – could invoke Article 24(1). That meant that the citizen had to be working, self-employed, studying, or have ‘sufficient resources’ not to become a ‘burden’ on the host state. The Court stated:

Dano has been followed in several cases since:

  • Alimanovic (Case C–67/14) [2016] 1 CMLR 29; [2016] QB 308: economically inactive Swedish nationals not entitled to social security benefits in Germany;
  • Mirga; Samin [2016] UKSC 1, [2016] 1 WLR 481 (UK Supreme Court): economically inactive Polish and Austrian nationals not entitled to social security benefits (income support and housing benefit, respectively) in the UK;
  • C.G. (Case C–709/20) [2022] 1 CMLR 26; [2021] 1 WLR 5919: economically inactive Dutch/Croatian national not necessarily entitled to social security benefits (Universal Credit) in the UK;
  • Fratila & Tanese [2021] UKSC 53 (UK Supreme Court): economically inactive Romanian nationals not entitled to Universal Credit in the UK.

The situation in C.G. was complicated by the fact that she was a single mother living in a women’s refuge with her two young children, following domestic violence perpetrated by her husband. The Court held that, where a Member State had granted an EU Citizen a right of residence on the basis of national law, and where that Citizen was ‘in a vulnerable situation’, the host State had a duty to ensure that that they could ‘live in dignified conditions’. The same applied to ‘children, who are particularly vulnerable’. It followed that the UK could refuse her application for Universal Credit but only after ascertaining that such refusal would not expose her and her children ‘to an actual and current risk of violation of their fundamental rights’, in particular the right to human dignity (Article 1 of the EU Charter).

Both Dano and Alimanovic were distinguished in J.D. (Case C–181/19) (2020) (unreported). The Court held that the benefit-claimant in that case, a Polish national who had worked in Germany but was subsequently (and temporarily) unemployed, was in a very different situation to the benefit-claimants in Dano and Alimanovic, neither of whom had ever worked in Germany. J.D.’s status as (i) an ex-worker and (ii) as the ‘primary carer’ of his two school-age daughters entitled him to continue to rely on the specific legislation conferring rights on workers, Regulation 492/2011, which included a right to claim ‘social advantages’ under the same conditions as the nationals of the host State. The J.D. case is discussed further in section 11.5.3, and Regulation 492/2011 is examined in more detail in Chapter 12 (see section 12.4.3).

Article 24(2) goes on to state that ‘the host Member State shall not be obliged to confer entitlement to social assistance during the first three months of residence’. This provision has been interpreted literally. Thus, in Garcia-Nieto & Others (Case C–299/14) [2016] 3 CMLR 5; [2016] 1 WLR 3089, the Court held that a provision of German legislation denying social security benefits during a citizen’s first three months of residence in Germany was perfectly lawful. This meant that the Garcia-Nieto family who left Spain and went to live in Germany in June 2012 would have to wait until September 2012 at the earliest until becoming eligible to even apply for German social security benefits.

11.5 The rights of a citizen’s family members and Directive 2004/38

11.5.1 The scope of the citizen’s family: Article 2

Directive 2004/38 extends the scope of the free movement provisions to a citizen’s ‘family members’. This is designed to further promote the free movement of citizens: there would clearly be a massive disincentive – financial as well as emotional – if citizens were not permitted by EU law to be accompanied by their families when working or studying abroad. Article 2(2) of the directive defines the citizen’s ‘family members’ as follows:

The Directive was adopted by the Council in April 2004 and its implementation date was 30 April 2006. However, it is very important to appreciate that this was not the first legislative definition of ‘family members’. Article 10 of Regulation 1612/68 contained the original definition and, although Article 10 has now been repealed, case law on the earlier Regulation continues to be relevant today.

‘Spouse’

The ECJ has held that ‘spouse’ means only marital relationships. In Netherlands v Reed (Case 59/85) [1986] ECR 1283 the ECJ stated:

This decision has attracted criticism (Clare McGlynn, ‘Families and the European Union Charter of Fundamental Rights’ (2001) 26 ELR 582) on the basis that it reinforces stereotypical assumptions about marriage. Notwithstanding this criticism, Reed was confirmed in Hadj Ahmed (Case C–45/12) (unreported), with the Court stating emphatically that ‘a mere cohabiting partner cannot be considered to be a “spouse”’.

Same-sex marriages

Do same-sex marriages confer ‘spousal’ status under Article 2(2)? Although several Member States recognise same-sex marriage (including Belgium, France, Germany, Ireland and Spain), many do not (including Bulgaria and Romania). In Coman (Case C–673/16) [2019] 1 WLR 425, the ECJ held that the term ‘spouse’ was ‘gender-neutral’ and may therefore cover a same-sex spouse. Where the host State did not recognise same-sex marriages, that State could not rely on its national law as justification for refusing to recognise a marriage concluded in another Member State that did recognise same-sex marriages. The Court said that this obligation did not ‘undermine the institution of marriage’ in the host State, nor did it require the host State to change its national law to allow marriages between persons of the same sex. This meant that Adrian Coman, a Romanian national, and his husband Robert Claibourn, a US national, who had married in Belgium, were entitled to return to Romania as Citizen and ‘spouse’, respectively. This was the case even though Romanian law did not recognise same-sex marriage.

This was confirmed in V.M.A. (Case C–490/20) [2022] 2 CMLR 22, in which V.M.A., a Bulgarian woman, had married K.D.K., a British woman, in Gibraltar. Thereafter, the two women lived together in Spain as Citizen and ‘spouse’, respectively (K.D.K. was no longer a Citizen as a result of Brexit so needed to invoke her spousal status to live in Spain.) In this case, the issue was not with Spanish law, which does recognise same-sex marriages, but Bulgarian law, which does not. This led the Bulgarian authorities to refuse to issue the couple’s daughter, S, with a birth certificate, because the official Bulgarian form required the name of the child’s ‘mother’ and ‘father’. V.M.A. argued that this refusal would make it ‘very difficult’ for S to get a Bulgarian passport, which in turn would hinder her rights of free movement. The ECJ held that the Bulgarian authorities were obliged to issue S with a Bulgarian passport under Article 4(3) of Directive 2004/38, whether or not she was entitled to a Bulgarian birth certificate under Bulgarian law. Moreover, doing so would not entail or require the making of any changes to Bulgarian law on the definition of marriage. (Article 4(3) states that ‘Member States shall, acting in accordance with their laws, issue to their own nationals, and renew, an identity card or passport stating their nationality’.)

Impact of separation

Separated, but not yet divorced, couples are still treated as spouses. In Diatta v Land Berlin (Case 267/83) [1985] ECR 567 the ECJ stated:

the marital relationship cannot be regarded as dissolved so long as it has not been terminated by the competent authority. It is not dissolved merely because the spouses live separately, even where they intend to divorce at a later date.

This is so even if they live apart. The ECJ has held that it is not necessary for spouses to cohabit (Diatta (1985)):

It is not for the immigration authorities to decide whether a reconciliation is possible. Moreover, if co-habitation of the spouses were a mandatory condition, the [citizen] could at any time cause the expulsion of his spouse by depriving her of a roof.

Diatta was followed in Iida (Case C–40/11) [2013] 1 CMLR 47, [2013] 2 WLR 788, involving a German woman and her Japanese husband. Although the couple were separated, they were not divorced and therefore Mr Iida was still classed as the ‘spouse’ of an EU citizen. Diatta was followed again in Ogieriakhi (Case C–244/13) [2015] 1 CMLR 14; [2014] 1 WLR 3823, involving a French woman (G) and her Nigerian husband (O) living in Ireland. The couple married in 1999 but separated in 2002, and O subsequently spent several years living with another woman. The couple eventually divorced in 2009. The ECJ held that O had retained his spousal status throughout this ten-year period of marriage, even though G and O had been living separately for seven years.

Location and timing of the marriage

The ECJ has held that it is immaterial where the marriage between citizen and spouse took place. In Metock & Others (Case C–127/08) [2008] ECR I–6241, the ECJ pointed out that there were no provisions in Directive 2004/38 stipulating ‘any requirements as to the place where the marriage of the Union citizen and the national of a non-member country is solemnised’. Moreover, it followed from that conclusion that there were no requirements that the marriage had to have taken place prior to the spouse entering the host Member State. This was despite the fact that Directive 2004/38, Article 3(1) states (emphasis added):

This Directive shall apply to all Union citizens who move to or reside in a Member State other than that of which they are a national, and to their family members as defined in Article 2(2) who accompany or join them.

In the Metock case itself, a number of non-EU nationals had entered Ireland, the host state (as asylum seekers) and each had subsequently met and then married a citizen of the Union who had arrived in that state after the asylum seeker. When the Irish government decided to commence deportation proceedings against the asylum seekers (who by this stage had had their asylum applications rejected), the ECJ ruled that the order in which the citizen and his or her spouse arrived in the host state was immaterial; what mattered was that the marriage was recognised as genuine.

The Court stated (emphasis added):

Although Article 3(1) proved to be no obstacle for the various spouses in Metock & Others, it was more problematic for the spouse in Iida (2013). The Court decided that Directive 2004/38 could not be invoked when the ‘spouse’ (and, by extension, all other family members) of the EU citizen was living in a state other than that in which the citizen was resident. This may at first appear harsh, but it is justifiable on the basis that EU law is designed to facilitate the free movement of persons from one Member State to another, which necessarily entails the right of EU citizens to take their family members with them when they move. There would be a serious obstacle to free movement otherwise.

Metock & Others was followed in Chenchooliah (Case C–94/18) [2020] 1 CMLR 19; [2020] 1 WLR 1801. Here, a woman from Mauritania went to Ireland as a student in 2005. Six years later, in 2011, she married a Portuguese national. That made her the ‘spouse’ of an EU Citizen, notwithstanding the fact that she had moved to Ireland several years before getting married.

The rule that ‘spouse’ means only marital relationships does not, however, mean that partners in other relationships do not have any rights (Reed (1986) – see below).

Implications of divorce/marriage annulment/termination of registered partnership

The question whether a divorced spouse remained entitled to residence in another EU Member State under EU law was left undecided in Diatta (1985), but in Baumbast (Case C–413/99) [2002] ECR I–7091 the ECJ made it clear that a divorcee cannot be regarded as a ‘spouse’. However, although a divorce terminates the ex-wife or ex-husband’s entitlement to residence as a ‘spouse’, that does not mean that they are no longer protected by EU law. Likewise with the situation of a former registered partner. Article 13 of Directive 2004/38 deals with these situations. Article 13(1) states that if an ex-wife, ex-husband or ex-partner are themselves an EU citizen then they can remain in the ‘host’ state, based on their own personal free movement entitlements. Article 13(2) deals with the more problematical situation where the ex-wife, ex-husband or ex-partner is a non-EU citizen (as would have been the case if the Senegalese wife in Diatta had actually divorced her husband).

Three cases have reached the ECJ on Article 13(2). In the first case, Singh & Others (Case C–218/14) [2016] 1 CMLR 12; [2016] QB 208, the Court decided that Article 13(2)(a) did not apply to cases in which the citizen had already left the host state before the initiation of divorce or annulment proceedings. The Court held that Article 13(2)(a) only allowed spouses or partners to retain a right of residence (under Article 7(2)) in the event of divorce of annulment, so that if that right had already been lost, prior to the commencement of those proceedings, then Article 13(2)(a) could not ‘revive’ it. In Singh & Others, this meant that the three former spouses (ex-husbands from Cameroon, Egypt and India) were not entitled to remain in the host state (Ireland) on the basis of Article 13(2)(a) because their EU citizen wives (from Germany, Latvia and Lithuania) had already left Ireland before the commencement of divorce proceedings. The wives’ departure from Ireland (whilst the couples were all still married) extinguished the (then) husbands’ residence rights under Article 7(2), which the subsequent divorce proceedings could not revive.

By way of contrast, in Baigazieva [2018] EWCA Civ 1088, the Court of Appeal held that a Kyrgyzstan national could rely on Article 13(2)(a) to retain her residence rights in the UK because she had initiated divorce proceedings from her husband (an EU Citizen of unspecified nationality) before he had left the UK and returned to his home State.

The second ECJ case, Ahmed (Case C–115/15) [2017] 1 CMLR 12; [2017] QB 109, involved a different provision, Article 13(2)(c), but a similar outcome to that in Singh & Others. The ECJ held that it was too late for a former spouse of an EU citizen to seek to invoke Article 13(2)(c) if divorce proceedings were commenced after the Citizen had already left the host state. However, the precedent established in Ahmed did not last long. In the third ECJ case, X v Belgium (Case C–930/19) [2022] 2 CMLR 1; [2022] 1 WLR 594, the Court took the highly unusual step of overruling itself. The Court in X said that the decision in Ahmed could provide the Citizen with a means of ‘exerting pressure’ on their spouse which would ‘clearly be contrary to the objective of ensuring the protection’ of a victim of domestic abuse, and thereby open that victim ‘to blackmail accompanied by threats of divorce or departure’. Therefore, divorce proceedings may be initiated after the departure of the Citizen from the host State in Article 13(2)(c) cases. However, in order to ensure legal certainty, a non-EU spouse ‘can rely on the retention of his or her right of residence … only in so far as those proceedings are initiated within a reasonable period following such departure’.

‘Marriages of convenience’

Although the ECJ has held that ‘spouse’ means ‘part of a married couple’ only, it subsequently had to qualify that statement by pointing out that ‘marriages of convenience’ will be regarded as an abuse of this situation. This point was made in Akrich (Case C–109/01) [2003] ECR I–9607, the facts of which will be examined in Chapter 12. The Court held that there would be an abuse of EU law if spousal rights to residence had been invoked in the context of ‘marriages of convenience’, entered into in order to ‘circumvent the national immigration provisions’.

In Rosa [2016] EWCA Civ 14, involving an alleged marriage of convenience between a Portuguese man and a Brazilian woman with the latter seeking a right of residence in the UK as a ‘spouse’, the Court of Appeal said that the ‘legal burden of proof on the issue of marriage of convenience lies throughout on the Secretary of State’, i.e. the Member State. This was confirmed by the UK Supreme Court in Sadovska [2017] UKSC 54; [2017] 1 WLR 2926, involving an alleged marriage of convenience between Violeta Sadovska, a Lithuanian national, and Saleem Malik, a Pakistani national.

The case of Kerim v Norway (Case E–1/20) [2021] 2 CMLR 19 involved an alleged marriage of convenience between a Romanian national and an Afghan national who had taken up residence in Norway. The EFTA Court (which has jurisdiction over the three Member States of the European Free Trade Association that are also members of the European Economic Area, i.e. Iceland, Norway and Liechtenstein) defined a ‘marriage of convenience’ as ‘an artificial construct’ on the basis of which a non-EEA national intended to improperly obtain a benefit under EU law. Thus, where there were reasonable doubts regarding whether a marriage was genuine, the national authorities (in this case, the Norwegian authorities) had to determine whether, had it not been for the ‘exclusive or essential purpose’ of improperly obtaining derived rights of free movement and residence for a non-EEA national (in this case, the Afghan ‘husband’), the marriage would not have been entered into by at least one of the spouses. The determination of the existence of a ‘marriage of convenience’ required an assessment of ‘a complex set of facts and circumstances, including the duration of the relationship, whether the parties resided together, had children together or shared parental responsibilities, and had serious long-term commitments together, such as financial commitments’. Moreover, since a genuine marriage was predicated upon the ‘good faith’ of both spouses, a statement from the Citizen (in this case, the Romanian ‘wife’) relating to the nature of the marriage and the purpose of entering into the marriage had to be considered and taken into account as part of that assessment. In Q v Norway (Case E–16/20) [2022] 2 CMLR 19, the EFTA Court considered another alleged ‘marriage of convenience’, this time between a Greek man (C) and a Peruvian woman (Q) who had also taken up residence in Norway. The Court stated that ‘A marriage of convenience is one in which the marriage was contracted in the absence of a genuine relationship between the parties and where the construct was purely artificial and entered into for the purposes of improperly obtaining a right under EEA law’. This case is considered in more detail below, in section 11.5.3.

‘Registered partners’

Some care is required here. Directive 2004/38 does not confer ‘family member’ status on ‘partners’ in the loose sense of a boyfriend or girlfriend (irrespective of the gender of the Citizen). Article 2(2)(b) of the directive only confers ‘family member’ status on ‘partners’:

  • who have ‘contracted a registered partnership, on the basis of the legislation of a Member State’; and
  • where ‘the legislation of the host Member State treats registered partnerships as equivalent to marriage’.

The reference to ‘the legislation of a Member State’ means that a partnership registered in, say, New Zealand would not count. Moreover, the host Member State must have legislation which ‘treats registered partnerships as equivalent to marriage’. Not all Member States have such legislation; for example, Bulgaria and Romania do not recognise civil partnerships.

However, unregistered partners – whatever their sexual orientation – may invoke Article 3(2)(b) instead. This refers to a ‘partner with whom the Union citizen has a durable relationship’ – clearly, a much wider concept. This will be examined in more detail below.

‘Descendants’

Article 2(2)(c) confers ‘family member’ status on any ‘direct descendants’ of the citizen, or of the citizen’s spouse or partner, provided that they are under 21 years of age or ‘dependent’. This most obviously applies to children (e.g. the daughter in Iida (2013)) and grandchildren (see for example Bigia & Others [2009] EWCA Civ 79, [2009] 2 CMLR 42, in which the Court of Appeal had no hesitation in holding that the Indian grandchildren of two Portuguese nationals living in the UK were entitled to the protection of Article 2(2)(c)).

As well as blood-relative children, this provision is designed to ensure that the citizen’s step-children are included in the family. The previous legislation, Article 10 of Regulation 1612/68, defined the ‘family’ in terms of a worker, his spouse and “their descendants”. This was more ambiguous about the position of step-children, and the ECJ was called upon to decide the point in Baumbast (2002), the facts of which were given above. The question was whether a citizen’s step-daughter (specifically, his wife’s daughter from a previous relationship) could be classed as one of ‘their descendants’. The Court – adopting a purposive rather than a literal interpretation of the legislation – held that she could.

In S.M. (Case C–129/18) [2019] 3 CMLR 16; [2019] 1 WLR 5505, the ECJ was asked whether an adopted child qualified as a ‘descendant’. The answer was yes.

‘Dependants’

Dependency needs to be established if the descendant is aged 21 or above. But what does it mean? The ECJ has held that the test of dependency is a question of fact, that is: does the Citizen provide financial support? In Lebon (1987), a case on Article 10 of Regulation 1612/68 involving a worker’s adult daughter, the ECJ held:

There is therefore no requirement that the citizen be under a legal obligation to provide support to his or her descendants. In Lebon (1987), the ECJ also ruled that a member of the citizen’s family does not cease to be dependent simply because they make a claim for a social welfare benefit. If it were otherwise, no member of a citizen’s family (apart from the spouse or registered partner and any children under 21) could ever make such a claim without taking themselves outside the scope of the family.

The ECJ looked again at the meaning of ‘dependency’ in the context of an adult descendant in Reyes (Case C–423/12) [2014] 2 CMLR 39; [2014] 3 WLR 1101. Coincidentally, this case also involved a 24-year-old daughter. The Court held that in order to qualify as a dependent descendant the ‘existence of a situation of real dependence’ must be established. That status must be the ‘result of a factual situation characterised by the fact that material support for that family member is provided’, either by an EU citizen or by his/her spouse or partner. In order to determine the existence of such dependence, the host Member State must assess whether, ‘having regard to [their] financial and social conditions’, the descendant was ‘not in a position to support’ themselves.

The Court added that the ‘need for material support must exist’, either in the descendant’s home country or in the country from which they came when they applied to join the citizen in the host state. However, there was no need to determine the reasons for that dependence. The fact that an EU citizen ‘regularly’ and ‘for a significant period’ paid a sum of money to the descendant in order for them to support themselves was evidence ‘to show that the descendant is in a real situation of dependence vis-à-vis that citizen’. The descendant could not be required to establish that they had tried to find work or obtain subsistence support or otherwise tried to support themselves. The descendant’s job prospects in the host state were irrelevant. The fact that a descendant was ‘deemed to be well placed to obtain employment’ because of their ‘personal circumstances’ such as youth, health and qualifications, and intended to start work in the host state, did not undermine their status as a dependant. Otherwise, a paradox would be created whereby any descendants aged 21 or over would lose the right to employment or self-employment (guaranteed by Article 23 of Directive 2004/38) on the basis that he or she was no longer a ‘family member’ precisely because they had good job prospects.

‘Ascendants’

Article 2(2)(d) confers ‘family member’ status on ‘dependent direct relatives in the ascending line’, whether that be the ‘line’ of the citizen or their spouse or partner (as the case may be). This provision is designed to allow citizen’s parents and grandparents, and in-laws, to claim ‘family member’ status. However, in a similar way to the situation of descendants who are aged 21 or over, ‘family member’ status is not automatic – the ascendant relative must be ‘dependent’. The situation is demonstrated by the case of Jia (Case C–1/05) [2007] ECR I–1.

One issue in particular was the appropriate test to be used to establish ‘dependency’. On this point, the ECJ stated:

Jia was applied by the Court of Appeal in Lim [2015] EWCA Civ 1383. The case involved a Malaysian national, Mrs Lim. She was both divorced and retired; owned a property in Malaysia worth £80,000; and had sufficient savings to meet her own needs. She also had two adult daughters, one of whom lived in the UK with her Finnish husband. Mrs Lim applied to join them in the UK as a dependent relative in the ascending line, but this was rejected. Although Mrs Lim’s daughter and sonin-law had regularly been sending her money, the evidence showed that she was in fact financially independent.

In Iida (2013), discussed above, the Court considered the possibility that Mr Iida, a Japanese national, might have a right of residence in Germany under Article 2(2) (d), on the basis that he was ‘dependent’ on his daughter. This was decisively rejected: it was, in fact, the ‘converse situation’, whereby she was still dependent on him, on the basis that he paid €300 per month towards her school fees, and not the other way around. The same decision was reached in Alokpa & Moudoulou (2017), Rendón Marín (2017) and V.I. (2022). In the former case, the Court held that Ms Alokpa, a Togolese national who lived in Luxembourg with her two young sons (both of whom were French nationals exercising their rights of residence) could not claim to be dependent on them. In the second case, the Court held that Mr Rendón Marín, a Colombian national who lived in Spain with his two young children (a boy with Spanish nationality and a girl with Polish nationality) could not claim to be dependent on either of them. And in the third case, Mrs V.I., a Pakistani national who lived in UK with her children (one of whom was an Irish national, having been born in Northern Ireland) could not claim to be dependent on her son. In all three cases, it was the same type of ‘converse situation’ as in Iida.

‘Irrespective of their nationality’

Under Directive 2004/38, the family members may be of any nationality. Among many cases demonstrating this point are Diatta (1985), where the spouse was from Senegal; Baumbast (2002), where the spouse and one child were Colombian nationals; Givane (2003), where the spouse and children were from India; Akrich (2003), where the spouse was from Morocco; Jia (2007), where the spouse and ascendant relatives were from China; and Iida (2013), where the spouse was Japanese. In Baumbast the ECJ confirmed:

No requirement that family members have lived in the citizen’s home state

In Akrich (2003), the ECJ had appeared to suggest that, in order to qualify for residence rights as a ‘family member’ – in that case, as a spouse – it was necessary that the person had been lawfully resident in the citizen’s home state. However, in Jia (2007), the ECJ stated that there was no such general requirement:

Hence, there was no legal obstacle to Mrs Jia, a Chinese national, taking up residence in Sweden as an ascendant relative of her son and daughter-in-law, despite the fact that she (Mrs Jia) had never set foot in Germany, her daughter-inlaw’s home state.

Situation when the citizen returns ‘home’

Consider this question: what would happen if, in the Jia case, the German daughter-in-law were to return to Germany, her ‘home’ state? Would her Chinese husband and mother-in-law be entitled to come to live in Germany with her? This question was raised in Eind (Case C–291/05) [2007] ECR I–10719, the Court answering ‘yes’.

Eind was decided under Regulation 1612/68 (subsequently repealed). The question whether the same entitlement arose under Directive 2004/38 arose in O & B (Case C–456/12) [2014] 3 CMLR 17; [2014] QB 1163. The Court held not. Article 3(1) of Directive 2004/38 defined the ‘beneficiaries’ of the rights conferred by it as ‘all Union citizens who move to or reside in a Member State other than that of which they are a national, and … their family members as defined in [Article 2(2)] who accompany or join them’ (emphasis added). It followed from a ‘literal, systematic and teleological’ interpretation of Directive 2004/38 that it did not establish a derived right of residence for non-EU family members in the Member State of which the citizen was a national. However, the Court went on to hold that, in principle, Article 21 TFEU instead conferred such a right, because otherwise, an obstacle to the citizen’s free movement rights would be created.

However, this was subject to the condition that ‘the residence of the Union citizen in the host Member State has been sufficiently genuine so as to enable that citizen to create or strengthen family life in that Member State’. Hence, Article 21 did not require that ‘every’ residence in the host state by a citizen accompanied by a non-EU family member necessarily conferred a derived right of residence on that family member in the citizen’s home state upon the citizen’s return to it. More specifically, residence by the citizen in the host state in accordance with Article 6 of Directive 2004/38 (i.e. for up to three months) was not evidence of an intention ‘to settle in the host Member State in a way which would be such as to create or strengthen family life in that Member State’. Short periods of residence such as weekends or holidays spent in the host state, even when considered cumulatively, remained within the scope of Article 6 and did not satisfy the condition of ‘genuine residence’. Conversely, residence in the host Member State pursuant to and in conformity with the conditions set out in Article 7(1) of the directive (i.e. for more than three months) was, in principle, evidence of ‘settling’ there and therefore of the citizen’s ‘genuine residence’ in the host state and went ‘hand in hand with creating and strengthening family life in that Member State’. This principle applied with particular strength where the citizen had acquired the right of permanent residence in the host state (i.e. after five years’ continuous legal residence) in accordance with Article 16(1) of the directive.

11.5.2 Other family members and partners in a durable relationship: Article 3(2)

Article 3(2) of Directive 2004/38 refers to some more categories of family members. It states that Member States are to ‘facilitate entry and residence for the following persons’:

  1. other family members, irrespective of their nationality, not falling under the definition in Article 2(2) who, in the country from which they have come, are dependants or members of the household of the Union citizen having the primary right of residence, or where serious health grounds strictly require the personal care of the family member by the Union citizen;
  2. the partner with whom the Union citizen has a durable relationship, duly attested.

‘Other family members’: Article 3(2)(a)

People such as a citizen’s siblings, uncles, aunts, cousins, nieces and nephews, who do not fall within the four categories of ‘family member’ above, are classed as ‘other family members’ (OFMs). However, it is not enough for, say, a citizen’s brother or nephew to claim OFM status based purely on their relationship. Article 3(2)(a) states that OFMs must, ‘in the country from which they have come’, be dependants or ‘members of the household of the Union Citizen’ or have ‘serious health grounds’ which ‘strictly require the personal care’ of the Union citizen. These concepts are all quite ambiguous, but some clarity has now been provided by the ECJ in Rahman (Case C–83/11) [2012] 3 CMLR 55, [2013] QB 249 and S.R.S. & A.A. (Case C–22/21) (2022), not yet reported.

‘The country from which they have come’

According to Rahman, this phrase is to be taken at face value, i.e. it simply refers to the country where the OFMs were living prior to joining the citizen in the host state. ‘The country’ does not have to be another EU Member State, and there is no requirement that the citizen had to have ever lived there. Hence, in Rahman, the two brothers-in-law of Roisin Rahman (an Irish national) were able to travel directly from their home country, Bangladesh, to join her and her husband (their brother) in the host state (the UK).

‘Dependants’

According to Rahman, the objective of Article 3(2)(a) is to ‘maintain the unity of the family in a broader sense’ by facilitating entry and residence for persons who are not included in the Article 2(2) definition of ‘family member’ but ‘who nevertheless maintain close and stable family ties with a Union citizen on account of specific factual circumstances, such as economic dependence’. In terms of timing, the Court stated that, whilst it was not essential that this relationship of dependency had to exist at the time when the Union citizen settled in the host state, it did have to do so at the time when the OFM applied to join him or her in that state. Whether or not it did so was a question of fact for the national court to decide.

‘Members of the household’

The ECJ considered the meaning of this phrase in S.R.S. & A.A. (2022). That Court held that Article 3(2)(a) required ‘more than mere cohabitation for reasons of pure convenience’. There had to be ‘a situation of genuine dependence’, based on ‘the existence of close and stable personal ties’ between the Citizen and the OFM. There had to be a ‘shared domestic life’ which had not been brought about with ‘the sole objective of obtaining entry into and residence in the host Member State’. The ‘degree of kinship’ between the Citizen and the OFM was ‘a factor to be taken into consideration’, but account should also be taken of ‘the closeness of the family relationship in question, reciprocity and the strength of the ties between those two persons’. The duration of the domestic life shared by the Citizen and the OFM was also ‘an important factor’, but this applied ‘irrespective of the date on which Union citizenship was acquired’.

Partners in a durable relationship (Article 3(2)(b))

A ‘partner in a durable relationship’ (PDR) is in the same position as an OFM. The concept of a ‘durable relationship’ is very ambiguous. For example, is ‘durability’ purely a question of longevity, in which case, how long? If there is more to durability than longevity, what else should be required – that the couple had moved in together? That they had children? What other criteria could be used to establish such a relationship? In Banger (Case C–89/17) [2019] 1 CMLR 6; [2019] 1 WLR 845, Rozanne Banger, a South African national, had lived with Philip Rado, a British national, in South Africa for two years before they both moved to the Netherlands. Three years’ later they moved again, this time to the UK. The ECJ was content to accept Ms Banger’s status as a PDR, simply observing that ‘Mr Rado and Ms Banger were neither married nor in a registered partnership but had been living together for several years’, and did not provide any guidance on the PDR concept.

However, there have been some UK cases which, if nothing else, indicate the sort of ‘relationship’ that domestic courts are likely to accept as being a PDR:

  • B & C v Home Secretary [2012] EWHC 226, [2012] 4 All ER 276: the High Court accepted that a Swedish man and his Bolivian girlfriend were in a ‘durable’ relationship – the couple had been together in London for over six years, they had lived together for most of that time, and they had a five-year-old daughter together.
  • Christy [2018] EWCA Civ 2378, [2019] 1 CMLR 30: the Court of Appeal accepted that Natasha Christy, an American national, had established a ‘durable’ relationship with a UK national, Mr Jones. They had met in London in 2011 and lived together in Poland before returning to the UK.
  • Aibangbee [2019] EWCA Civ 339: the Court of Appeal accepted that Jefferey Aibangbee, a Liberian national, was in a durable relationship with a Czech national, the couple having lived together in the UK since 2008.

Rights of OFMs and partners in a durable relationship

Neither OFMs nor partners in a durable relationship have a guaranteed right of entry and residence, and nor do they have any rights to employment. Article 3(2) simply states that ‘the host Member State shall … facilitate entry and residence’, although it adds that the state concerned must ‘undertake an extensive examination of the personal circumstances and shall justify any denial of entry or residence to these people’. In Rahman, the ECJ explained that Article 3(2) did not ‘oblige’ the Member States to accord a right of entry and residence to OFMs. However, because of the words ‘shall … facilitate’, it was clear that OFMs had ‘a certain advantage’ compared to people who had no connection to an EU citizen at all, because they were at least entitled to have their application for entry and residence carefully considered by the host state.

11.5.3 The concept of ‘primary carer’

In addition to the above legislative definitions, the ECJ has added a further category of family member: the ‘primary carer’. In Baumbast and R v Home Secretary (2002) the ECJ invented this concept as a means of conferring a continued right of residence on two mothers who would otherwise have faced deportation from the UK (and possible separation from their children):

  • Baumbast v Home Secretary. Mrs Baumbast, a Colombian national, faced being deported because her German husband had left the UK to work, initially in China and subsequently in Lesotho in southern Africa. She was therefore no longer the spouse of an EU citizen in the UK.
  • R v Home Secretary. Mrs R, an American national, faced deportation following her divorce from her French husband because she was no longer the spouse of an EU citizen.

However, the ECJ held that both women retained rights of residence as ‘primary carer’ of their children, who were still being educated in British schools. The ‘primary carer’ doctrine is technically based on Article 10 of Regulation 492/2011, which confers on all children of Union citizens who have worked in another Member State the right to be educated in that state. (Article 10 is examined in more detail in Chapter 12.) In Baumbast and R, the Court stated:

The judgment in Baumbast and R did leave some questions unanswered. In particular, for how long does ‘primary carer’ status continue? That question has now been answered by the ECJ in Teixeira (Case C–480/08) [2010] ECR I–1107. The ECJ held that ‘in principle’ children who have reached adulthood are assumed not to require parental care, but that it was possible for the ‘primary carer’ to be needed beyond that age to help their son or daughter complete their education.

The Court stated:

The ‘primary carer’ therefore enjoys a right of residence as long as they have a ‘child’ in education who needs parental support. In Alarape & Tijani (Case C–529/11) [2013] 3 CMLR 38; [2013] 1 WLR 2883, the ECJ followed Teixeira and confirmed that a primary carer’s derived right of residence potentially lasted until the ‘child’ completed his or her education, even if by that point the ‘child’ had reached adulthood. The Court added that the determination of whether an adult ‘child’ continued to need the presence and care of his or her parent in order to pursue and complete their education was a question of fact to be resolved by the national court. This meant that Olaitan Alarape, a Nigerian national, was potentially entitled to remain in the UK in order to care for her son, Olukayode Tijani, even though he was by this point in time a 24-year-old PhD student at Edinburgh University. (Olukayode qualified as the ‘child’ of a ‘worker’ because Olaitan had been married to a French national whilst the latter was working in the UK – a worker’s stepchild can rely on Article 10.) The ECJ did state that the circumstances and features which the national court may take into account included, inter alia, the age of the child, whether the child resided in the family home and whether the child needed financial and/or emotional support from the parent in order to be able to continue and to complete his or her education.

There are no other conditions or restrictions on the primary carer’s rights of residence, a point that was made very clearly in Ibrahim (Case C–310/08) [2010] ECR I–1065, decided on the same day as Teixeria.

The Court stated:

This proposition established in Ibrahim, that children in education, plus their primary carer, were entitled to residency in the host State using Article 10 of Regulation 492/211, without any of them needing to satisfy the residency conditions in Article 7 of Directive 2004/38, was confirmed in J.D. (Case C–181/19) (2020) (unreported). The case involved an unemployed Polish national living in Germany with his two school-age daughters. Crucially, he had previously worked in Germany which meant that his daughters were covered by Article 10. The Court held that

Article 10 grants to a child, in parallel with the right that child has to access to education, an independent right of residence that does not depend on the fact that the parent or parents who care for the child should continue to have the status of migrant worker in the host Member State.

In Czop & Punakova (Cases C–147, 148/11) [2013] Imm AR 104, the ECJ refused to expand the scope of the ‘primary carer’ concept to include the children of the self-employed. This meant that two otherwise essentially identical cases were decided differently:

  • Lucja Czop, a Polish national, had come to live in the UK in 2002, initially as a student, and subsequently in self-employment. Her eldest child was born in Poland, and she had three other children, who were all born in the UK to the same father, a self-employed Polish national called Krzyzowski. The Court held that Lucja did not qualify as a primary carer because, despite the fact that her children were studying in the UK, none of them were the children of a ‘worker’.
  • Margita Punakova, a Czech national, came to live in the UK in 2001. She subsequently gave birth to three children; their father was a Lithuanian national called Buklierius, who was intermittently employed in the UK. The Court held that Margita’s children were the children of a (former) worker, and therefore Margita was entitled to remain in the UK as their primary carer.

The Court explained:

In Hadj Ahmed (Case C–45/12) (2013) (unreported) the Court refused to extend the primary carer concept to a case in which the putative primary carer (an Algerian national) had never been married to the citizen (a French national working in Belgium) and the child in question (also Algerian) was the former’s daughter from a previous relationship. As the daughter was not the ‘child of a national of a Member State’, Article 10 did not apply to her, and hence the primary carer concept did not apply to her mother either.

In Q v Norway (Case E–16/20) (2022), a Greek man (C), his Peruvian wife (Q) and her son from a previous relationship (A) went to live in Norway. C worked for a period of time but then the couple separated and he returned to Greece. By this time, A was attending school in Norway. The Court held that – in principle at least – A was entitled to rely on Article 10 as the ‘child’ of a person who had worked in the host State (i.e., C); A’s mother, Q, was entitled to live with him as his ‘primary carer’. (Although A was C’s stepson, A could still be regarded as C’s ‘child’ following Alarape & Tijani (2013).) However, the Norwegian authorities alleged that C and Q’s marriage was not genuine; rather, it had been one of ‘convenience’. The Court stated that ‘EEA States may refuse, terminate or withdraw any right conferred by EEA law in the case of abuse or fraud, such as a marriage of convenience’. This meant that if the marriage was found to be one of ‘convenience’ then both Q and A faced having their rights to live in Norway ‘terminated’.

11.5.4 Rights to employment: Article 23

It is particularly important to bear in mind that this right is conferred on all family members, ‘irrespective of nationality’, so that even non-EU nationals can work legally in the EU. Article 23 replaced Article 11 of Regulation 1612/68, which has now been repealed. However, the two leading cases on Article 11 remain relevant as a guide to the interpretation of Article 23. Both of the cases involve family members – spouses – who were non-EU nationals.

Article 11 was applied in Gül (Case 131/85) [1986] ECR 1573, involving a Cypriot national, a doctor, who had been refused the right to work in Germany, despite being the spouse of a UK national who was employed there. The ECJ held that such a refusal was in breach of Regulation 1612/68, and emphasised that the authorities of the host state were obliged to ‘treat the spouse in a non-discriminatory fashion’ in so far as access to employment was concerned.

It should be noted that Cyprus became a Member State of the EU in May 2004, which would have entitled Dr Gül to rely upon EU law in his own right. Nevertheless, the principle of law established remains applicable to other non-EU family members.

In Mattern & Cikotic (Case C–10/05) [2006] ECR I–3145, the ECJ was again asked to rule on Article 11 of Regulation 1612/68. The case involved a Luxembourg national, Cynthia Mattern, who was married to a ‘Yugoslav’ national, Hajrudin Cikotic. Ms Mattern had obtained work in Belgium, so clearly Mr Cikotic would have been entitled to work there, applying the precedent from Gül. However, the issue was whether he was entitled to work in a different state – specifically, Ms Mattern’s home state of Luxembourg (the couple lived near the border between those two countries). The ECJ refused, stating:

11.6 Limitations on free movement – the derogations in Directive 2004/38

The right of free movement for citizens and their family members is not an ‘absolute’ right. Specifically, the right is subject to ‘limitations justified on grounds of public policy, public security or public health’. Further guidance is given in Article 27 of Directive 2004/38:

This provision largely replaces Articles 1–3 of Directive 64/221, which only dealt with workers and their family members, and which has now been repealed. However, many of the cases decided by the ECJ under the old Directive have continuing relevance today, given that both Directives use very similar – in some cases, identical – terms. Indeed, some of the jurisprudence of the ECJ has been adopted into the new Directive. Compare, for example, Article 27(2), above, with the quote from the case of R v Bouchereau (1978), below.

11.6.1 Scope of the derogations

The ECJ had handed down a number of important judgments on the scope of Directive 64/221. Although that legislation has now been repealed, many of those judgments are still relevant as a guide to the scope of the derogations under Directive 2004/38.

Directive 2004/38 only applies to restrictions placed on the movement of ‘Union citizens and their family members’

This point is made explicitly in Article 27(1) but had already been stated by the ECJ in Commission v Spain (Case C–503/03) [2006] ECR I–1097, a case involving Directive 64/221. The ECJ stated that EU law:

enables Member States to prohibit nationals of other Member States or their spouses who are nationals of third countries from entering their territory on grounds of ‘public policy’ or ‘public security’. The [Union’s] legislature has nevertheless made reliance by the Member States on such grounds subject to strict limits.

Implicit in this judgment is a principle that Directive 2004/38 does not apply to any case only involving a non-EU national (or, as the Court describes them, ‘nationals of third countries’). This can be seen in the decision in Dem’Yanenko (Case C–45/03) (2005), unreported, where the ECJ declined jurisdiction in a case involving the deportation of a Ukrainian national from Italy. Two cases in which a citizen’s non-EU spouse claimed protection under Article 27 are Rendón Marín (2017) and C.S. (2017), which will be discussed in section 11.6.8.

EU nationals do not need to have been lawfully resident in another Member State in order to be protected

This point was made in another case involving Directive 64/221, Commission v Netherlands (Case C–50/06) [2006] ECR I–1097, where the ECJ stated:

The fact that Member States cannot deport their own nationals does not prevent them from deporting other states’ nationals

This point was emphasised in Pereira Roque (Case C–171/96) [1998] ECR I–4607, where the ECJ stated (emphasis added):

11.6.2 ‘Public policy’

The situations when ‘public policy’ measures may be invoked are not necessarily concerned with criminal activities (see Van Duyn v Home Office (Case 41/74) [1974] ECR 1337 – permission to enter the UK refused to a member of the Church of Scientology). The ECJ has given guidance on what sort of behaviour will justify public policy measures. In R v Bouchereau (Case 30/77) [1978] ECR 1999, the ECJ held:

In Adoui and Cornuaille (1982) the ECJ noted that prostitution could amount to a threat to public policy. In Calfa (Case C–348/96) [1999] ECR I–11, confirmed in Orfanopoulos and Oliveri (Case C–482/01) [2004] ECR I–5257, the ECJ stated that ‘the use of drugs constitutes a danger for society such as to justify special measures against foreign nationals who contravene its laws on drugs, in order to maintain public order’. There is no requirement that the activity be criminal according to the law of the Member State (Van Duyn (1974)). However, it is important that the Member State concerned is regulating the conduct of which the citizen is accused among its own nationals. This point was made in Adoui and Cornuaille (1982).

The Court stated:

This point was reiterated in Jany and Others (Case C–268/99) [2001] ECR I–8615, involving six women who were refused residence permits despite working as self-employed prostitutes in Amsterdam. The ECJ stated that EU law did not ‘impose on Member States a uniform scale of values as regards the assessment of conduct which may be considered to be contrary to public policy’, before repeating its judgment in Adoui (1982).

In Oteiza Olazabal (Case C–100/01) [2002] ECR I–10981, the ECJ authorised the imposition of ‘public policy’ measures involving an EU citizen’s residence rights in specific areas of another Member State. The ECJ stated:

It seems that the ECJ reached its decision because otherwise the complainant (a Spanish national convicted of terrorism offences in France) would have to be deported from France. This would not be in keeping with the fundamental requirement of proportionality.

In Rutili (Case 36/75) [1975] ECR 1219, the ECJ explained that the ‘genuine and serious threat’ test was designed to ensure compliance with Articles 8–11 of the European Convention on Human Rights (ECHR), that restrictions were available in the interests of national security or public safety only when necessary for the protection of those interests in a democratic society.

11.6.3 ‘Public security’

This is reserved for serious crimes and subversive, anti-state activities, e.g. terrorism and espionage. Public security was specifically referred to in Oteiza Olazabal (2002), involving a member of ETA (Euskadi Ta Askatasuna [Basque Homeland and Freedom]). Formed in 1959, the organisation seeks autonomy for the Basque area of northern Spain. The ECJ stated:

An important case on the scope of the ‘public security’ derogation is Tsakouridis (Case C–145/09) [2010] ECR I–11979. The Court was asked whether involvement in criminal activity (in this case, drug trafficking) could be classified as a public security (rather than public policy) matter. It answered in the affirmative:

In P.I. (Case C–348/09) [2012] 3 CMLR 13, [2012] QB 799, the ECJ was asked whether the public security derogation could be applied to someone convicted of child sex offences. Again, the Court responded in the affirmative, holding that it was open to the Member States to regard certain criminal offences, such as the sexual exploitation of children, as constituting a particularly serious threat to one of the fundamental interests of society, which might pose a direct threat to the calm and physical security of the population and thus be covered by the concept of public security ‘as long as the manner in which such offences were committed discloses particularly serious characteristics’. Whether or not that was the case was for the national courts to decide.

11.6.4 ‘Public health’

Article 29 of Directive 2004/38 provides as follows:

There is no ECJ case law (yet) to provide any guidance on phrases such as ‘diseases with epidemic potential’ or ‘other infectious diseases or contagious parasitic diseases’. At the start of the Covid-19 pandemic, in March 2020, all of the EU’s Member States adopted various forms of border controls to try to stop the spread of the virus. Responding to these unprecedented events, the European Commission issued a series of guidelines, ‘COVID-19 Guidelines for Border Management Measures to Protect Health and Ensure the Availability of Goods and Essential Services’. In them, the Commission stated that ‘maintaining the functioning of the Single Market is key’ and that, whilst the imposition of (temporary) border controls were permissible, ‘Member States must always admit their own citizens and residents, and facilitate transit of other EU citizens and residents that are returning home’. On the subject of border controls, the Commission guidelines stated, amongst other things, that:

  • 18. Member States may reintroduce temporary border controls at internal borders if justified for reasons of public policy or internal security. In an extremely critical situation, a Member State can identify a need to reintroduce border controls as a reaction to the risk posed by a contagious disease.
  • 19. Such controls should be applied in a proportionate manner and with due regard to the health of the individuals concerned. Persons who are clearly sick should not be refused entry but should have access to appropriate health care.…
  • 21. For EU citizens, the safeguards laid down in Directive 2004/38 must be guaranteed. In particular, non-discrimination between Member States’ own nationals and resident EU-citizens must be ensured. A Member State must not deny entry to EU citizens or third-country nationals residing on its territory and must facilitate transit of other EU citizens and residents that are returning home. Member States can, however, take appropriate measures such as requiring persons entering their territory to undergo self-isolation or similar measures upon return from an area affected by Covid-19 provided they impose the same requirements on their own nationals.…
  • 23. Member States should permit and facilitate the crossing of frontier workers, in particular but not only those working in the health care and food sector, and other essential services (e.g. child care, elderly care, critical staff for utilities) to ensure continued professional activity.…
  • 25. Member States, and in particular neighbouring Member States, should closely cooperate and coordinate at EU level to ensure effectiveness and proportionality of the measures taken.

11.6.5 ‘Proportionality’

Article 27(2) states that ‘Measures taken on grounds of public policy or public security shall comply with the principle of proportionality.’ An example of this can be seen in the case of Oulane (Case C–215/03) [2005] ECR I–1215, where a French national had been arrested behaving suspiciously at Rotterdam railway station in the Netherlands, in an area closed off to the public. He was unable to produce any ID, nor could he give a residence address and had no luggage. He was detained and, a week later, he was deported. Subsequently he brought a legal challenge to the deportation. The ECJ decided that deportation in such circumstances would be disproportionate. The Court stated:

In Tsakouridis (2010), the Court provided further guidance on the meaning of the ‘proportionality’ requirement. The Court stated:

The case of Byankov (Case C–249/11) [2013] 1 CMLR 15, [2013] 2 WLR 293 illustrates the ‘proportionality’ test. The case involved Bulgarian legislation which prevented those who owed a ‘considerable’ debt to another person from leaving the country. This was held to be disproportionate (and therefore contrary to Article 27(1)) because the prohibition was ‘absolute’, that is, it was ‘not coupled with any exceptions, temporal limitation’ – i.e. it could last indefinitely – ‘or possibility of regular review of the factual and legal circumstances underpinning it’.

11.6.6 ‘Personal conduct’

Article 27(2) of the Directive states: ‘measures taken on grounds of public policy or public security shall be based exclusively on the personal conduct of the individual concerned’. Applying this, the ECJ has held that Member States may not justify deportation of an individual on the ground that it will serve as an example to others. In Bonsignore (Case 67/74) [1975] ECR 297, the ECJ emphasised how public policy and/or public security measures ‘cannot be justified on grounds extraneous to the individual’. In the same case the ECJ held that ‘The concept of “personal conduct” expresses the requirement that a deportation order may only be made for breaches of the peace and public security which might be committed by the individual affected’. In Calfa (1999), the complainant faced automatic deportation from Greece, having committed a relatively minor drugs offence. The ECJ held that deportation was not permissible without taking account of the individual’s personal conduct.

The Court stated:

This seems correct. After all, there is a huge difference between, on one hand, someone who is caught with a substantial amount of a highly dangerous drug such as heroin which they clearly intend to sell and, on the other hand, someone who has only a small amount of a recreational drug for her own personal use. Yet, under Greek law, all drug offenders were subject to the same blanket rule: automatic lifetime expulsion.

A similar provision of national legislation was examined by the ECJ in Oliveri (2004). The case concerned the Ausländergesetz (German law on aliens), which provided for mandatory deportation for foreign nationals in certain circumstances, one being where a foreign national committed a drugs offence or a public order offence and was sentenced to a term of imprisonment (para 47). The ECJ stated that such mandatory deportation would be incompatible with the Directive.

The Court also held that a practice adopted in Germany, whereby courts were prohibited, when reviewing the legality of national authorities’ deportation orders, from taking into account any subsequent factual developments, constituted a breach of the personal conduct criterion.

The Court stated:

The Court dealt with a similar issue in Orfanopoulos (2004). This case involved another provision of the Ausländergesetz, under which certain foreign nationals have ‘special protection’, including those living in a family relationship with a German national (para 48). This protection means that deportation is only available on ‘serious’ grounds of public security and policy. When a foreign national entitled to special protection commits an offence that would normally lead to mandatory deportation, such as those in para 47, he can still be deported – not automatically, but ‘as a general rule’. In effect, there is a presumption in favour of deportation. The Court held that the presumption in para 48 also breached the personal conduct requirement of the directive. The Court held that it was essential in every case to take into account a number of factors:

  • the nature and seriousness of the offences committed;
  • the length of residence in the host Member State;
  • the period which has elapsed since the commission of the offence;
  • the family circumstances of the person concerned;
  • the seriousness of the difficulties which the spouse and any of their children risk facing in the country of origin of the person concerned.

The Court stated:

Membership of organisations and the ‘personal conduct’ rule

According to Van Duyn (1974), membership of an organisation may constitute ‘personal conduct’. Thus, members of organisations known to pose a threat to public policy or security (most obviously terrorist groups) can be made subject to deportation orders or refusals of entry purely on the basis of their membership. However, the ECJ distinguished between past and present membership.

The Court stated (emphasis added):

This decision is highly significant, as it allows Member States to expel EU nationals who are known to be a present member of a terrorist organisation, without having to prove that the individual has committed any specific offence. For example, in Gallagher (Case C–175/94) [1995] ECR I–4253, the UK authorities expelled a known IRA member. Although there was evidence that he had committed a firearms offence, under the Van Duyn (1974) ruling this would not have been necessary. Establishing his membership of the IRA alone would have sufficed. Another case example is Oteiza Olazabal (2002), discussed above. Again, he had been convicted by a French court of a terrorist offence but, under Van Duyn (1974), restrictions could have been placed on his free movement simply by establishing his membership of the terrorist group ETA.

Van Duyn has application beyond terrorist organisations. In R v Kraus [1982] Crim LR 468, the Court of Appeal authorised the removal of a German national from the UK, partly on the basis of criminal activity, and partly on the basis of his membership of the Nazi party. Watkins LJ said that ‘We are entitled to take notice … that members of parties such as the Nazi party are notorious for their unswerving allegiance to the objectionable cause they seek to serve and for the ruthlessness with which they seek to achieve their aims’. More recently, in Wahl v Iceland (Case E–15/12) [2014] 1 CMLR 29, the EFTA Court applied Van Duyn. The Court held that the Icelandic authorities were entitled to refuse entry on public policy grounds to a Norwegian national who was known to be a member of the Hell’s Angels motorcycle gang, described in the judgment as an ‘international motorcycle club associated with organised crime’.

11.6.7 Excluded situations

Directive 2004/38 provides that measures taken on the grounds of public policy, security or health will not be justified in two situations.

‘Economic ends’: Article 27(1)

This would mean, for example, that a Member State could not deny entry to an EU citizen from another state on the ground that there was high unemployment in the state. The meaning of ‘economic ends’ was considered in Aladzhov (Case C–434/10) [2011] ECR I–11659. A Bulgarian businessman had been prevented from leaving Bulgaria until he paid around €22,000 in back taxes. He challenged this, arguing that it would severely impact on his business interests which required a lot of foreign travel. The ECJ held that non-payment of tax may fall within the scope of ‘public policy’. Moreover, since the purpose of taxes was to ensure the ‘funding of actions of the Member State on the basis of the choices which are the expression of, inter alia, its general policy in economic and social matters’, measures adopted by a Member State in order to ensure that they were paid could not be said to have been adopted exclusively to serve ‘economic ends’.

‘Previous criminal convictions’: Article 27(2)

Previous criminal convictions shall not ‘in themselves’ constitute grounds for the taking of such measures. So what significance do they have? In R v Bouchereau (1978) the ECJ emphasised that the individual had to represent a ‘present’ threat.

The Court offered the following guidance:

The decision in R v Bouchereau (1978) suggested that a single conviction of a serious offence could be enough to justify deportation. The propositions established in Bouchereau were applied and developed in K (Case C–331/16) [2018] 3 CMLR 26; [2019] 1 WLR 1877. Here, the ECJ stated that where restrictions on a Citizen’s free movement and residence rights were imposed based on past conduct amounting to alleged criminal offences, the ‘public policy’ assessment had to take into account ‘the nature and gravity of the crimes or acts that that individual is alleged to have committed, the degree of his individual involvement in them and the possible existence of grounds for excluding criminal liability such as duress or self-defence’. Such an examination was all the more necessary in a situation where the person concerned had not actually been convicted of the crimes or acts in question. Finally, although the Court confirmed that the Bouchereau test primarily looks to the Citizen’s future conduct (‘propensity’) rather than past acts, the Court also stated that where past acts were of ‘exceptional gravity’ then – even after a relatively long period of time – the person responsible for them could be classified as a ‘persistent’, i.e. a ‘present’, threat.

11.6.8 Procedural safeguards

The need to examine the individual’s circumstances: Article 28(1)

Article 28(1) of Directive 2004/38 provides as follows:

This list of ‘considerations’ in Article 28(1) echoes the list provided by the ECJ in Orfanopoulos (2004). More recently, in both Rendón Marín (2017) and C.S. (2017), the ECJ (without referring to Article 28(1)), emphasised the importance of taking into account the effect of deportation of any children who might be affected. The Court stated that an expulsion decision involving a person who is the sole carer of a child (or children) can result ‘only from a specific assessment by the national court of all the current and relevant circumstances of the case [including] the child’s best interests…. That assessment must therefore take account [of] the age of the child at issue and his state of health, as well as his economic and family situation’.

In Dumliauskas & Others [2015] EWCA Civ 145, the Court of Appeal identified another potentially relevant ‘consideration’: the (relative) prospects of rehabilitation in the individual’s home state and the host state. Sir Stanley Burnton said that:

Special protection for those with a right of permanent residence: Article 28(2)

Article 28(2) applies to anyone with a right of permanent residence, meaning those who have been resident for at least five years in the host state. It states:

The ECJ has yet to consider what is meant by ‘serious’ policy or security grounds, but the Court of Appeal has done so. In Bulale [2008] EWCA Civ 806, [2009] QB 536, the Court held that there was a distinction to be drawn between crimes of dishonesty (such as theft) and crimes of violence (including robbery). Only the latter category of criminal offending would cross the ‘serious’ threshold, opening up the possibility of the authorities taking an expulsion decision against those with permanent residency. The case itself involved a Dutch national (B) who had resided in the UK for over five years, thereby acquiring a right of permanent residence. However, he had also committed a number of criminal offences, including robbery. The Court of Appeal held that this entitled the Home Secretary to deport B, on the basis that his involvement in violent crime constituted ‘serious’ public policy grounds.

Extra protection for those with ten years’ residence and minors: Article 28(3)

Article 28(3) applies to those with at least ten years’ residence in the host state, and to minors.

The leading case on Article 28(3) is Tsakouridis (2010).

‘Imperative grounds’

In both Tsakouridis and P.I. (2012), the Court was asked whether public security (rather than public policy) applied. The reason was that both cases involved EU citizens (from Greece and Italy, respectively) who had lived in the host Member State (Germany in both cases, coincidentally) for over ten years and hence could only be deported on ‘imperative grounds of public security’. As well as explaining the scope of ‘public security’ and the ‘proportionality’ requirement, in Tsakouridis the Court offered some guidance on the meaning of the phrase ‘imperative grounds’, when it said (emphasis added):

‘Resided … for the previous ten years’: Article 28(3)(a)

The Court in Tsakouridis also addressed this phrase. The question arose because the Greek national in that case had travelled extensively between Greece and Germany during the ten years preceding his conviction for drug trafficking, which called into question his entitlement to rely upon Article 28(3)(a). The Court stated: Judgment

In Onuekwere (2014), discussed in section 11.4.9, the Court held that time spent in prison did not count as ‘residence’ for the purpose of acquiring a right of permanent residence under Article 16. In M.G. (Case C–400/12) [2014] 2 CMLR 40; [2014] 1 WLR 2441, the ECJ reached exactly the same conclusion with respect to Article 28(3)(a). Moreover, the Court held that the ten-year period of residence necessary for the grant of the enhanced protection provided for in Article 28(3)(a) must be calculated by counting backwards from the date of the deportation order, not by counting forwards from the date of the individual’s arrival in the host state. The Court also ruled that, in principle, ‘such periods interrupt the continuity of the period of residence for the purposes of that provision’.

The effect of this is that if a deportation order is issued against an individual who is in, or has only recently been released from, prison, it will be very difficult, if not impossible, for them to invoke Article 28(3)(a). In B & Vomero (Case C–316/16; Case C–424/16) [2018] 3 CMLR 24; [2019] QB 126, the ECJ was asked to clarify the operation of Article 28(3), particularly as some Member States made expulsion decisions immediately after conviction (and hence before prison) whereas other Member States waited and made their decision during or even after prison. This could potentially lead to ‘arbitrary or unfair results’, depending on when any expulsion decision was adopted. The Court began by holding that the ‘enhanced protection’ provided in Article 28(3) was available only to Citizens who satisfied the eligibility criterion for the protection referred to in Article 28(2), namely a right of permanent residence under Article 16(1). The Court then confirmed the principles established in M.G., i.e. that 10-year period of residence provided for in Article 28(3) must be (i) calculated by counting back from the date of the expulsion decision and (ii) continuous (at least, in principle). This meant that periods of imprisonment interrupted the continuity of the period of residence, for the purpose of Article 28(3) (again, in principle).

However, the fact that the person concerned was imprisoned could not be regarded as automatically breaking any ‘integrative links’ that that person had previously forged with that host State and the ‘continuity’ of their residence in that State. The more those ‘integrative links’ with that State were ‘solid’ from a ‘social, cultural and family perspective’, to the point where the person concerned was ‘genuinely rooted in the society of that State’, the lower the probability that a period of detention could have resulted in those links being broken and, consequently, a discontinuity of the 10-year period of residence. Other factors included the nature of the offence, the circumstances in which that offence was committed, and ‘all the relevant factors as regards the behaviour of the person concerned during the period of imprisonment’. This meant that where a Citizen who had already resided in the host State for a period of at least 10 years was imprisoned, and an expulsion decision was made during or after prison – such that the term of imprisonment formed part of the 10-year period counting back from the adoption of that decision – it did not automatically entail a discontinuity of that 10-year period.

Children: Article 28(3)(b)

Article 28(3)(b) applies to cases involving children, ruling out deportation except where it would be in their ‘best interests’. This is an intriguing concept, and it will be interesting to see what guidance the ECJ offers as to when it will be in a child’s ‘best interests’ to be deported.

Giving reasons: Article 30

Article 30(1) states: ‘the persons concerned shall be notified in writing of any decision taken under Article 27(1), in such a way that they are able to comprehend its content and the implications for them’. This does not mean that the notification has to be in the individual’s own language; they just have to be able to ‘comprehend’ it. In Petrea (Case C–184/16) [2018] 1 CMLR 42; [2018] 1 WLR 2237, the Court held that Article 30(1) does not require ‘that the removal order is to be translated into the language of the person concerned, but requires by contrast that the Member States take the necessary measures to ensure that the latter understands the content and implications of that decision’.

Article 30(2) adds that the person concerned should be informed of the grounds of public policy, public security or public health upon which the decision taken in his case is based, unless this is contrary to the interests of the security of the state involved. The authorities must give the person a precise and comprehensive statement of the reasons for its decision (Rutili (1975)). Without adequate reasons, the person cannot prepare a full defence.

There is an apparent contradiction inherent in Article 30(2) in public security cases, requiring as it does Member States’ obligation to inform the person concerned of the public security grounds on which any restriction on free movement is based – unless the interests of state security take precedence. The ECJ was asked to clarify Article 30(2) in Z.Z. (Case C–300/11) [2013] 3 CMLR 46; [2013] QB 1136. The Court ruled that Article 30(2) ‘must be interpreted strictly, but without depriving it of its effectiveness’. This meant that Member States were permitted not to disclose ‘certain information’ to the person concerned in the light of ‘overriding considerations connected with State security’. More specifically, the Court acknowledged that full disclosure of evidence in public security cases may ‘endanger the life, health or freedom of persons or reveal the methods of investigation specifically used by the national security authorities and thus seriously impede, or even prevent, future performance of the tasks of those authorities’.

However, the Court also held that the Member States had the task of proving that state security would in fact be ‘compromised’ by precise and full disclosure. It followed that there was ‘no presumption that the reasons invoked by a national authority exist and are valid’. At the very least, the person concerned must be informed of ‘the essence’ of the grounds on which a decision refusing entry taken under Article 27(1) was based. The Court stressed that ‘the necessary protection of State security cannot have the effect of denying the person concerned his right to be heard’.

Appeals and reviews: Article 31

Article 31 provides for appeals against, and reviews of, measures taken by Member States to restrict the free movement of other states’ nationals. Article 31(1) provides:

This replaces Article 8 of Directive 64/221, which has been repealed, although some of the case law on that provision is of continuing relevance, for example Royer (1976) and Pecastaing (1981).

In Royer, the ECJ was asked whether a deportation order could be executed immediately, or whether such an order became effective only after all national judicial remedies had been exhausted. The ECJ stated: Judgment

In Pecastaing (Case 98/79) [1981] ECR 691, the ECJ stated:

A Member State cannot … render the right of appeal for persons covered by the directive conditional on particular requirements as to form or procedure which are less favourable than those pertaining to remedies available to nationals in respect of acts of the administration.

Duration of exclusion orders: Article 32

Article 32 deals with the duration of any exclusion order made by a Member State. Article 32(1) provides:

Expulsion as a penalty or legal consequence: Article 33

Article 33(1) provides that Member States may not make expulsion an automatic consequence of a person from another Member State committing a criminal offence in their territory. This is consistent with pre-existing ECJ case law on the ‘personal conduct’ criterion, examined above, in particular Calfa (1999) and Orfanopoulos & Oliveri (2004). In that sense Article 33(1) simply confirms existing law and practice. Article 33(2) provides: The separation in time between a deportation order being made and its enforcement, as contemplated by Article 33(2), is likely to occur if an order is made when a convicted criminal is serving a prison sentence, with the order to take effect on release. This situation can be illustrated using the facts of Santillo (Case 131/79) [1980] ECR 1585. This case actually involved a different issue – the question for the ECJ was whether a trial judge’s recommendation to deport could be regarded as valid some four-and-a-half years later (the ECJ ruled that this was a matter for national courts to decide). Were the facts of this case to recur now, then there would be no need to have recourse to Article 33(2) as the date when the Home Secretary’s deportation order was made (September 1978) was only seven months before the date when the order was enforced (April 1979).

11.6.9 Application to Member States’ own nationals

The vast majority of the above cases in this section involve restrictions imposed by Member States on other states’ nationals. However a Member State can also impose restrictions on its own nationals. An example is Jipa (Case C–33/07) [2008] ECR I–5157.

The Court stated:

There have since been other cases before the ECJ involving restrictions on Member States’ own nationals, including Aladzhov (2012) and Byankov (2013), discussed earlier, both of which involved Bulgarian nationals who were prevented from leaving Bulgaria on public policy grounds linked to unpaid taxes and an unpaid debt, respectively.

Even before Jipa, the Court of Appeal had already addressed this point, in the context of football banning orders, in Gough and Others [2001] 4 All ER 289. These are orders which may be imposed under the UK’s Football Spectators Act 1989 (as amended) on English football hooligans in order to prevent them travelling to watch matches in Europe involving the England national side or English clubs. When banning orders were imposed on four individuals, they challenged the legality of the orders under EU law (and the European Convention on Human Rights). Laws LJ dismissed the challenges, holding that: ‘in a proper case a Member State may be justified on public policy grounds in preventing a citizen of the Union from leaving its shores’. A more recent example of a restriction being imposed by the UK on a British national is X.H. (2017), who had his passport cancelled because the government suspected that he intended to travel to Syria to participate in terrorist activities.

11.7 Citizens’ political rights: Articles 22–25 TFEU

Article 22(1) TFEU states that every citizen of the Union, if residing in an EU Member State of which he or she is not a national, has the right to both vote in, and even to stand as a candidate at, ‘municipal elections’ of the host state ‘under the same conditions as nationals of that State’. This would mean, for example, that any French or German national living in Ireland would be entitled to vote in any Irish municipal elections provided that they were at least 18 years of age (the minimum Irish voting age). They could also stand for election if they were at least 18 years of age (the minimum age for Irish candidates) and, if elected, could then serve provided they were not disqualified (Irish law disqualifies certain people from holding political office, such as members of the judiciary, police force or armed forces, or anyone currently sentenced to more than 6 months in prison).

Article 22(2) adds:

every citizen of the Union residing in a Member State of which he is not a national shall have the right to vote and to stand as a candidate in elections to the European Parliament in the Member State in which he resides, under the same conditions as nationals of that State.

Article 22(2) was examined by the ECJ in Eman & Sevinger (Case C–300/04) [2006] ECR I–8055. The case involved two Dutch nationals who were resident in Aruba, a Caribbean island some 15 miles off the coast of Venezuela. Aruba is classed as an overseas territory of the Netherlands. However, the Dutch authorities had refused to register them for the European Parliament elections in 2004 because of provisions in Dutch law which conferred the voting franchise on all Dutch nationals resident in the Netherlands (but excluding the Dutch Antilles and Aruba), plus other Member States’ nationals resident in the Netherlands. The complainants contested that this was a breach of their rights under Article 22(2).

However, the ECJ rejected their claim on this basis. Although the Court accepted that EU citizens resident in one of the ‘overseas countries and territories’ (OCTs) could rely on EU law, it held that Article 22(2) did not apply to ‘a citizen of the Union residing in an OCT who wishes to exercise his right to vote in the Member State of which he is a national’. The Court stated that, as a general principle, there was nothing in the present state of EU law which prevented Member States from insisting on residency as a criterion in conferring the franchise on voters and also for establishing the right to stand for election. Here the Court noted that the European Court of Human Rights had earlier held, in Melnychenkov v Ukraine (2004), that the obligation to be resident within national territory in order to be able to vote is a requirement which is not, in itself, unreasonable or arbitrary and which can be justified on several grounds. The ECJ concluded: ‘the criterion linked to residence does not appear, in principle, to be inappropriate to determine who has the right to vote and to stand as a candidate in elections to the European Parliament’.

Article 23 TFEU confers certain diplomatic rights on citizens of the Union. Specifically, citizens have the right to protection by the diplomatic or consular authorities of other Member States if they are in a non-EU country and there is no diplomatic or consular office from that person’s own state. For example, a Latvian national – and hence also a citizen of the Union – would be able to rely upon protection from (say) a Dutch or Spanish embassy in any country in the world if there happened to be no Latvian diplomatic presence in that country.

Finally, Article 24 TFEU allows citizens of the Union to:

  • petition the European Parliament;
  • apply to the Ombudsman established under Article 228 TFEU; and
  • write to ‘any of the institutions’ (namely the European Parliament, the European Commission, the Council, the European Council, the ECJ, the European Central Bank and the Court of Auditors) in any of the ‘authentic’ languages mentioned in Article 55(1) TEU and to receive a reply in the same language.

Applying the Law

Adèle is due to travel from her home in Paris to Ireland next month (June) to take up a teaching position at University College Dublin. Adèle has a medical condition which means that she cannot drive. She has applied to the Irish Department of Social Protection for a disability benefit which would cover the cost of her daily commute, but has been told she would need to wait until December on the basis that benefit claimants need to have lived in Ireland for at least six months.

Adèle currently lives with her girlfriend, Besjana, who is Albanian. The couple met three years ago and have lived together in Paris since January 2020. Besjana has read on the Irish Immigration Service website that she would only be allowed to stay in Ireland for a maximum of 90 days on a tourist visa, and would not be able to take up paid employment. Adèle and Besjana are considering their options, which include hurriedly entering into a “civil solidarity pact” under French law or waiting and getting married in Ireland. Ideally, Besjana would love to set up her own craft shop in Dublin. Assuming that she is allowed to settle in Dublin, Besjana would like to bring her elderly, widowed grandmother, Rovena, to live with them. Rovena lives alone in Tirana, Albania.

Adele and Besjana now seek your advice as to their rights under EU law. Advise:

  1. Adèle of her right to claim the disability benefit in Ireland;
  2. Adèle whether Besjana can join her in Dublin;
  3. Besjana whether she would be allowed to set up her own business in Dublin in the event that she is allowed to enter the country;
  4. Besjana whether she can bring Rovena to Ireland to live with them.

SUMMARY

  • Article 20 TFEU establishes ‘Citizenship of the Union’ for every person holding the nationality of a Member State, including dual nationals (Micheletti, Collins). The acquisition of nationality is a matter of national law (Micheletti, Zhu & Chen).
  • The loss of nationality is also a matter of national law but if it also involves the loss of EU Citizenship, then it may breach Article 20 TFEU (Rottmann, Tjebbes & Others, J.Y.).
  • The citizen need not be economically active. Citizenship rights may be invoked by the unemployed (Martínez Sala); students (Grzelczyk, D’Hoop, Bidar, Förster); the retired (Pusa); and those incapable of working for health reasons (Tas-Hagen).
  • The age of the citizen is immaterial. Children can invoke citizenship status (Zhu & Chen).
  • Citizens have ‘the right to move and reside freely within the territory of the Member States’ (Article 21 TFEU), which is directly effective (Baumbast & R). Citizens may invoke Article 21 TFEU, in combination with Article 18 TFEU, in order to challenge national legislation which discriminates (directly or indirectly) against nationals of other Member States.
  • Citizens may also invoke Article 21 TFEU in order to challenge national legislation which places them at a ‘disadvantage’ following the exercise of free movement rights (Pusa, Tas-Hagen).
  • Citizens must not become ‘an unreasonable burden’ on the host state (Grzelczyk). But this does not mean that citizens cannot claim financial benefits (Martínez Sala, D’Hoop, Grzelczyk, Collins, Bidar). The key word is ‘unreasonable’.
  • Certain restrictions on citizens’ rights are justifiable if based on ‘objective considerations of public interest’ do not directly discriminate on grounds of nationality, are suitable and proportionate. One ‘objective consideration’ is the need to ensure that there is a ‘real link’ between the claimant and the State (D’Hoop).
  • Citizens have a range of political rights, e.g. to participate in elections (Article 22 TFEU).
  • Under Directive 2004/38, citizens have a right of ‘exit’ from their home state (Article 4); a right of ‘entry’ into the host state (Article 5); a right of residence in the host state for up to three months without any conditions or formalities (Article 6). For longer residence periods, the citizen should either be working, self-employed, financially independent or a student (Article 7).
  • Citizens may acquire a right to remain permanently in the host state after five years’ continuous residence (Article 16). Time spent in prison does not count as ‘residence’ (Onuekwere). A right of permanent residence may also be acquired on reaching retirement age (after three years’ residence), or if forced to retire from work as a result of permanent incapacity (Article 17).
  • Citizens are entitled to equal treatment with the nationals of that Member State ‘within the scope of the Treaty’ (Article 24). Economically inactive citizens cannot rely on Article 24 to claim benefits (Dano, Alimanovic) unless denying benefits would be a ‘violation of their fundamental rights’ (C.G.).
  • Directive 2004/38 confers rights on a citizen’s ‘family members’, ‘irrespective of their nationality’:
  • ‘Spouse’ refers to genuine marital relationships only (Reed, Akrich). It includes same-sex marriages (Coman, V.M.A.). It is immaterial where or when the marriage was solemnised (Metock & Others). Marital status continues after separation (Diatta, Iida, Ogieriakhi). The spouse is not obliged to remain in the same accommodation as the citizen (Diatta). Divorce terminates the spouse’s status (Baumbast & R) but under Directive 2004/38, Article 13, ex-spouses retain residence rights in some situations (Singh & Others, X v Belgium).
  • ‘Registered partners’, but only if the legislation of the host Member State treats registered partnerships as equivalent to marriage.
  • ‘Descendants’ of the citizen/spouse/partner, if under 21 or ‘dependent’. ‘Dependence’ is a factual issue (Lebon, Reyes).
  • ‘Dependent relatives in the ascending line’ of the citizen/spouse/partner. The ‘need for material support must exist in the State of origin’ (Jia).
  • Directive 2004/38, Article 3, confers limited rights of entry and residence on ‘other family members’, irrespective of their nationality, if ‘dependants’ (Rahman) or ‘members of the household’ of the citizen (S.R.S. & A.A.), or where ‘serious health grounds strictly require … personal care’, and on ‘the partner with whom the Union citizen has a durable relationship, duly attested’ (Banger).
  • Family members have rights to take up ‘employment or self-employment’ (Directive 2004/38, Article 23), but only in the same Member State as the citizen (Gül, Mattern & Cikotic), and to ‘enjoy equal treatment’ with nationals of the host state (Directive 2004/38, Article 24).
  • Under Directive 2004/38, Member States may impose ‘measures’ which restrict free movement on grounds of public policy, public security or public health, such as refusal of entry (Van Duyn, K), refusal of exit (Jipa), expulsion (R v Bouchereau), and territorial restrictions (Rutili).
  • Although Member States cannot refuse entry to/expel from national territory their own nationals, they are not prevented from refusing entry to/expelling other states’ nationals (Pereira Roque).
  • The concept of ‘public policy’ must ‘be interpreted strictly, so that its scope cannot be determined unilaterally by each Member State’ (Rutili). ‘Public policy’ measures require a ‘genuine, present and sufficiently serious threat affecting one of the fundamental interests of society’ (Directive 2004/38, Article 27(2)). States must be prepared to impose ‘repressive’ measures on their own nationals before invoking ‘public policy’ measures (Adoui & Cornuaille, Jany).
  • Public security has been invoked in cases involving terrorism (Oteiza Olazabal), organised crime (Tsakouridis) and child sex offences (P.I.).
  • Public policy or security measures must be based ‘exclusively’ on the individual’s ‘personal conduct’ and not on considerations of general prevention (Directive 2004/38, Article 27(2); Bonsignore, Calfa, Orfanopoulos & Oliveri). But present membership of an organisation may constitute ‘personal conduct’ (Van Duyn, Wahl).
  • Previous criminal convictions must not ‘in themselves’ constitute grounds for the taking of measures (Article 27(2)) but may do so if they provide evidence of a ‘present threat’ (R v Bouchereau, K).
  • Measures taken on grounds of public policy or security must comply with the principle of ‘proportionality’ (Article 27(2); Oulane, Tsakouridis).
  • Before seeking expulsion, Member States must take into account how long the individual concerned has resided on its territory, his/her age, state of health, family and economic situation, social and cultural integration into the host Member State, and the extent of his/her links with the country of origin (Directive 2004/38, Article 28(1)). The impact of deportation on any children is particularly important (Rendón Marín, C.S.).
  • For individuals with a permanent right of residence, expulsion is only possible on ‘serious’ grounds of public policy or security (Article 28(2)). Minors, or someone with ten years’ residence, may only be expelled ‘on imperative grounds of public security’ (Article 28(3), Tsakouridis, P.I.). Time spent in prison does not count as ‘residence’ (M.G.).
  • States may take public health measures against those suffering from ‘diseases with epidemic potential’ as well as ‘infectious diseases or contagious parasitic diseases’ (Directive 2004/38, Article 29).
  • All decisions require notification ‘in writing’ (Directive 2004/38, Article 30; Z.Z.). The individual must be able to appeal and/or seek judicial review of any decision taken against them (Article 31(1)). Persons excluded on public policy or security grounds may apply to have it lifted after three years (Article 32).

Further reading

Articles

  • Azoulai, L and Coutts, S, ‘Restricting Union Citizens’ Residence Rights on Grounds of Public Security’ (2013) 50 CML Rev 553.
  • Currie, S, ‘Accelerated Justice or a Step Too Far? Residence Rights of non-EU Family Members and the Court’s Ruling in Metock’ (2009) 34 EL Rev 310.
  • Dautricourt, C and Thomas, S, ‘Reverse Discrimination and Free Movement of Persons Under Community Law: All for Ulysses, Nothing for Penelope’ (2009) 34 EL Rev 433.
  • Haag, M, ‘The Coup de Grace to the Union Citizen’s Right to Equal Treatment’ (2022) 59 CML Rev 1081.
  • Hylten-Cavallius, K, ‘Who Cares? Caregivers’ Derived Residence Rights from Children in EU Free Movement Law’ (2020) 57 CML Rev 399.
  • Kochenov, D, ‘The Right to Have What Rights? EU Citizenship in Need of Clarification’ (2013) 19 ELJ 502.
  • Łazowski, A, ‘Darling You Are Not Going Anywhere: The Right to Exit and Restrictions in EU Law’ (2015) 40 EL Rev 877.
  • O’Brien, C, ‘Real Links, Abstract Rights and False Alarms: The Relationship between the ECJ’s “Real Link” Case Law and National Solidarity’ (2008) 33 EL Rev 643.
  • O’Brien, C, ‘The Great EU Citizenship Illusion Exposed: Equal Treatment Rights Evaporate for the Vulnerable’ (2021) 46 EL Rev 801.
  • Reynolds, S, ‘Exploring the “Intrinsic Connection” between Free Movement and the Genuine Enjoyment Test: Reflections on EU Citizenship after Iida’ (2013) 38 EL Rev 376.
  • Rijpma, J, ‘You Gotta Let Love Move’ (2019) 15 ECL Rev 324.
  • Ristuccia, F, ‘“’Cause Tramps Like Us, Baby We Were Born to Run”: Untangling the Effects of the Expulsion of “Undesired” Union Citizens’ (2022) 59 CML Rev 889.
  • Solanke, I, ‘Another Type of “Other” in EU Law?’ (2013) 76 MLR 383.
  • Tryfonidou, A, ‘The ECJ Recognises the Right of Rainbow Families to Move Freely between EU Member States’ (2022) 47 EL Rev 534.
  • Van Eijken, H, ‘Tjebbes in Wonderland: on European Citizenship, Nationality and Fundamental Rights’ (2019) 15 ECL Rev 714.

12 The free movement of workers

DOI: 10.4324/9781003218562-12

12.1 The objectives of Article 45 TFEU

The objectives of Article 45 are, broadly speaking, twofold:

12.1.1 To allow for workers to move from one EU Member State to another for the purposes of employment

This benefits both individual workers and their employers. Individual workers benefit because they can move from areas of high unemployment and/or low wages to areas of low unemployment and/or higher wages. This is true whether the workers are skilled or not, and whether they possess professional qualifications or not. Readers of this book who are familiar with the 1980s TV programme Auf Wiedersehen, Pet will recall how the plot of the original series involved British manual workers (bricklayers, plasterers, etc.) moving from the UK (where their skills were not in high demand) to Germany, where their skills were put to good use working on building sites.

A real-life example of people exploiting EU rules on the free movement of workers is provided by the huge numbers of professional footballers from countries across the EU who have gone to play in other Member States, particularly since the Bosman ruling of December 1995, which applied Article 45 to professional sport. These players typically move to bigger, more successful clubs in more lucrative leagues where they can exploit their earning potential and boost their prospects of wining trophies and/or earn international recognition. Notable examples include Robert Lewandowski, who left his native Poland in 2010 to play for Borussia Dortmund and Bayern Munich in Germany and currently plays for Barcelona in Spain; Cristiano Ronaldo, who left his native Portugal to play for Manchester United in 2003 before moving on to Real Madrid in Spain and then Juventus in Italy before returning to United in 2021; and Antoine Griezmann, a French footballer who has spent his entire professional club career in Spain playing for Real Sociedad, Atlético Madrid and Barcelona. All of these players are ‘workers’ who invoked their rights under Article 45 to go from their home countries to play – or, rather, work – in another.

Employers also benefit because they have a greater choice of potential workers to choose from. This may mean they can employ workers with better qualifications or greater experience than they otherwise would be able to if restricted to workers from the same Member State. Workers from states where there are relatively low average wages may be prepared to move to other states to take on unskilled work, or work anti-social hours (which national workers may be unwilling to do) and still receive better salaries than they would had they stayed at home. An influx of foreign workers can also help resolve ‘skills shortages’. At the time of the EU’s expansion in May 2004 it was widely reported in the British media that highly skilled engineers and scientists from central and eastern Europe (Poland in particular) were coming to the UK to work. This benefited the individual workers (average wages in the UK being much higher than those in Poland), and their employers, who were struggling to find suitably qualified British engineers and scientists. As far as the UK is concerned, Brexit has brought an end to the free movement of workers but the principle is very much alive and well across the EU.

12.1.2 To prohibit discrimination on grounds of nationality against workers who have moved

Article 45(2) prohibits Member States, through legislation, and employers, through their terms and conditions of employment, from discriminating against workers who have moved under Article 45(1). After all, there would be a massive disincentive to move if employers were free openly to discriminate against migrant workers. The dual objectives of Article 45 were neatly summarised by the ECJ in the case of Lyyski (Case C–40/05) [2007] ECR I–99, when the ECJ stated:

12.2 The scope of Article 45

Article 45(1) provides that ‘Freedom of movement for workers shall be secured within the Union’. This means that workers are (subject to very narrowly defined derogations) free to move from one Member State to another without restriction. The ECJ offered some guidance on this in Graf (Case C–190/98) [2000] ECR I–493:

The Court stated:

The principle established in Graf was illustrated in Casteels v British Airways (Case C–379/09) [2011] ECR I–1379. Here, a Collective Agreement between British Airways (BA) and the relevant trade union in Germany stated, inter alia, that employees who transferred from a division of BA in Germany to a division of BA in another Member State (but not those who transferred within Germany) were deemed to have left the company ‘voluntarily’, which had adverse consequences when it came to calculating their length of service with the company, and ultimately their pension entitlement. The Court held that this provision placed workers

who exercise their right to free movement at a disadvantage [which was] liable to dissuade workers from leaving their employer’s establishment in one Member State in order to take up a position with an establishment of that same employer in another Member State.

This dissuasive effect was enough to bring the provision within the scope of Article 45.

Article 45(2) states that, having moved, workers are entitled not to be discriminated against in terms of nationality. This provision is directly effective and can be enforced in national courts in order to challenge discriminatory national legislation, or to challenge discriminatory employment practices. Article 45(2) even applies to private employers, as Angonese (Case C–281/98) [2000] ECR I–4139 demonstrates:

The ECJ has held that Article 45(2) can also be invoked by employers. In Innovative Technology Center (Case C–208/05) [2007] ECR I–181, the Court decided that the right of workers to be engaged and employed without discrimination necessarily entailed, as a corollary, an employer’s entitlement to engage them without discrimination. That right also entailed, as a further corollary, the right of intermediaries, such as recruitment agencies, to assist work-seekers in finding employment. Hence, provisions of German law, under which the Federal Employment Agency was obliged to pay recruitment agencies for finding work for unemployed people – but only if the employment was based in Germany – constituted a prima facie infringement of Article 45(2).

The principle that employers can invoke Article 45 as well as employees was seen again in Caves Krier Frères (Case C–379/11) [2013] 2 CMLR 14. A wine producer/merchant in Luxembourg invoked Article 45 to challenge a rule of national law that offered a financial incentive to Luxembourg-based companies to recruit older workers, but only if they were resident in Luxembourg. This was held to restrict the free movement of workers based in other Member States (who would be less likely to be offered employment by companies in Luxembourg), and by corollary it restricted the freedom of Luxembourg companies to employ migrant workers.

Article 45(2) applies throughout the territories of the Member States and even extends beyond the EU itself, to workers who, although situated outside the geographical scope of the EU, have an employment relationship with their employer which is founded within one of the Member States or where the relationship retained a ‘sufficiently close link’ with the EU. In SARL Prodest (Case 237/83) [1984] ECR 3153 the ECJ ruled:

12.3 The definition of ‘worker’

The definition of ‘worker’ has the same meaning in all EU Member States because, otherwise, Member States could define ‘worker’ in such a way as to prevent the access of migrant workers (Levin (Case 53/81) [1982] ECR 1035). However, the word is not defined in any EU legislation. The ECJ has therefore been called upon in several cases to explain the meaning and scope of the ‘worker’ concept. In Lawrie-Blum (Case 66/85) [1986] ECR 2121, the ECJ noted that the free movement of workers was a fundamental principle, and it therefore had to be defined widely. It then stated:

The employer need not necessarily be a company of any particular nationality. EU nationals employed by international organisations based in another EU Member State are entitled to the protection afforded by Article 45:

  • Echternach (Case 389/87) [1989] ECR 723 – German national employed by the European Space Agency in Belgium;
  • Schmid (Case C–310/91) [1993] ECR I–3011 – German national employed by Eurocontrol (the European Organisation for the Safety of Air Navigation) in Belgium;
  • Ferlini (Case C–411/98) [2000] ECR I–8081 – Italian national employed at the European Commission in Luxembourg;
  • My (Case C–293/03) [2004] ECR I–12013 – Italian national employed at the Council of the European Union in Brussels, Belgium;
  • Öberg (Case C–185/04) [2006] ECR I–1453 – Swedish national employed at the European Court of Justice in Luxembourg;
  • Rockler (Case C–137/04) [2006] ECR I–1441 – Swedish national employed at the European Commission in Brussels, Belgium;
  • Alevizos (Case C–392/05) [2007] ECR I–3505 – Greek national and member of the Greek Air Force, seconded to the North Atlantic Treaty Organisation (NATO) in Italy;
  • Gardella (Case C–233/12) [2013] ICR D34 – Italian national employed by the European Patent Office (EPO) in Germany;
  • Adrien & Others (Case C–466/15) and P.F. & Q.G. (Case C–27/20) (both unreported) – French nationals employed at the Court of Justice in Luxembourg.

12.3.1 Part-time and low-paid employees

The ECJ has been very flexible here. Some activity of an economic nature, even if it is poorly paid, qualifies an individual as a ‘worker’. On the question of whether part-time workers are covered, the ECJ has stated (Levin 1982):

The ECJ had consistently refused to lay down a rule setting any sort of quantitative threshold of work, either in terms of the number of hours per week or in terms of a minimum salary. Rather, what matters to the ECJ is that a basic qualitative threshold of work has been met.

The fact that a worker has to supplement his income in order to subsist is also irrelevant – thereby narrowing what the ECJ suggested in Levin (1982): that employment must be ‘effective’. This principle is clearly illustrated in Kempf (Case 139/85) [1986] ECR 1741.

The Court stated:

Although the ECJ has been generous with its definition of ‘worker’, the principle will not be taken to extremes. Where a person’s activities can be described as ‘marginal and ancillary’, they will not be classed as a ‘worker’. The ECJ has subsequently stated that the duration of the activity performed is relevant in determining whether activities amount to genuine employment or are simply marginal and ancillary (Raulin (Case C–357/89) [1992] ECR I–1027).

In Ninni-Orasche (Case C–413/01) [2003] ECR I–13187, the ECJ explicitly refused to hold that the fact that an Italian national had been employed for two-and-a-half months over a period of three years in Austria automatically excluded her from the scope of Article 45. Ultimately, whether or not she was a ‘worker’ was a question of fact for the national court, applying the ‘effective and genuine activities’ test. The ECJ stated:

Two cases have raised questions regarding whether or not work was ‘genuine’. In Bettray (Case 344/87) [1989] ECR 1621, involving a German national who was living in the Netherlands, the ECJ held that activities described as ‘social employment’ which were carried out as part of a state-sponsored drug rehabilitation programme, did not constitute genuine employment. However, in Trojani (Case C–456/02) [2004] ECR I–7573, the ECJ took a more generous view with what was described as ‘a personal socio-occupational reintegration programme’. Although the ECJ distinguished Bettray (1989), it conceded that the ultimate decision whether or not a person qualified as a worker was a question for the national court.

The Court stated:

It should be remembered that the ultimate decision as to whether or not a person qualifies as a ‘worker’ is a matter for the national court, applying the principles in cases such as Lawrie-Blum and Levin. For example, in Barry v Southwark LBC [2008] EWCA Civ 1440, [2009] 2 CMLR 11, the Court of Appeal held that a Dutch national who had worked in the UK as a security guard for about nine months followed by two weeks as a steward during the Wimbledon tennis championship qualified as a ‘worker’, on the basis that his employment was ‘effective and genuine’. Conversely, in B & Others v Home Secretary [2012] EWCA Civ 1015, [2012] 3 CMLR 43, the same court held that an Italian national who had lived in the UK for over ten years but had only worked (as a cleaner) for a total of 80 hours (and had been motivated to take this employment purely in order to maintain his benefits) was not a ‘worker’.

12.3.2 Trainees

Students undertaking vocational training are ‘workers’. In Lawrie-Blum (1986), involving a British woman who had applied to become a trainee teacher in Germany, the ECJ held:

Lawrie-Blum (1986) was followed in Bernini (Case C–3/90) [1992] ECR I–1071, involving an Italian woman working as a trainee in a furniture factory in the Netherlands. In Kranemann (Case C–109/04) [2005] ECR I–2421, the ECJ was asked whether a trainee lawyer qualified as a ‘worker’. The case involved a German national who had completed his legal education in Germany and was spending part of his legal training with a law firm in London. The ECJ had no doubt that the trainee was a ‘worker’:

12.3.3 Work-seekers

Could Article 45 be interpreted so as to give work-seekers protection? A literal reading of Article 45(3) would seem to indicate not, as it states that the freedom of movement for workers includes the right ‘to accept offers of employment actually made’. However, the ECJ very rarely adopts a literal reading of EU legislation, preferring to take a purposive approach (favouring any interpretation which promotes free movement over one which inhibits it). Hence, in Antonissen (Case C–292/89) [1991] ECR I–745, the ECJ held that work-seekers did have certain rights:

The ECJ added that migrant nationals seeking employment in other EU Member States had to be able to ‘appraise themselves … of offers of employment’ and to take the ‘necessary steps in order to be engaged’. Work-seekers could not, therefore, be deported as long as they could provide evidence that they were (a) continuing to seek employment and (b) had genuine chances of becoming employed.

In G.M.A. (Case C–710/19) [2021] 2 CMLR 13, the ECJ clarified the situation regarding work-seekers’ residence rights in the host State. The Court held that:

  1. The host State is required to allow work-seekers a ‘reasonable period of time’ (RPT) to ‘apprise themselves of offers of employment corresponding to their occupational qualifications and to take, where appropriate, the necessary steps in order to be engaged’.
  2. The RPT starts to run from the time when the work-seeker registers as a job-seeker in the host State.
  3. A period of six months from that date did not appear, ‘in principle, to be insufficient’.
  4. A work-seeker may be required to actively seek employment during the RPT. The work-seeker must ‘continue’ to seek employment after expiry of that period in order to retain residence rights in the host State.
  5. During the RPT, the host State cannot require the work-seeker to demonstrate the existence of a ‘genuine chance of being engaged’.
  6. It was only after the RPT had elapsed that the work-seeker may be required to provide such evidence.

Work-seekers and social benefits

In Collins (Case C–138/02) [2004] ECR I–2703, the ECJ held that a work-seeker could also claim entitlement to certain social benefits, such as job-seeker’s allowance. That does not guarantee they will receive those benefits, however. In Collins, the issue was whether a particular benefit for unemployed people in the UK could be made subject to a requirement of ‘habitual residence’. The ECJ held that this requirement was potentially indirectly discriminatory against foreign nationals, on the basis that UK nationals would be more likely to satisfy it, but it was nevertheless justifiable. The Court stated that it ‘may be regarded as legitimate for a Member State to grant such an allowance only after it has been possible to establish that a genuine link exists between the person seeking work and the employment market of that State’.

However, the ‘habitual residence’ test was subject to the ‘proportionality’ principle. The ECJ held that its application by the national authorities ‘must rest on clear criteria known in advance and provision must be made for the possibility of a means of redress of a judicial nature’. Finally, any period of residence laid down in the national rules ‘must not exceed what is necessary in order for the national authorities to be able to satisfy themselves that the person concerned is genuinely seeking work in the employment market of the host Member State’.

Collins was followed in Vatsouras & Koupatantze (Cases C–22, 23/08) [2009] ECR I–4585, involving two Greek nationals who had briefly worked in Germany before being made unemployed. They both applied for, but were denied, social benefits. They challenged the refusals. The Court held that, in general, work-seekers fell within the scope of Article 45 and therefore enjoyed the right to equal treatment laid down in Article 45(2). In particular, this included the right to claim financial benefits – but only if those benefits were ‘intended to facilitate access to the labour market’. In Alhashem [2016] EWCA Civ 395, the Court of Appeal held that the claimant (a Dutch national looking for work in the UK) was not entitled to Employment and Support Allowance. This was ‘primarily provided for those who cannot work or who are on the borderlines due to some disability or past episode in their lives’; it was therefore not a benefit which was ‘intended to facilitate access to the labour market’.

Moreover, the ECJ confirmed in Vatsouras & Koupatantze that it was legitimate for a Member State to grant such benefits only after the work-seeker had established a ‘real link’ with the labour market of that state (which was for the competent national authorities and, where appropriate, the national courts to establish).

12.3.4 The previously employed

The ECJ has held that persons who have lost their job but are capable of taking another are still within the definition of ‘worker’. In Leclere and Deaconescu (Case C–43/99) [2001] ECR I–4265 the ECJ stated:

This is a remarkable statement – a person who is seeking work is, clearly, not actually working. However, the Court says that such a person is nevertheless to be ‘classified’ as a worker. It appears that this statement only applies to those persons who have at one time been workers for the purposes of Article 45. Those people who have never worked – such as the claimants in Antonissen (1991) and Collins (2004) – remain in the pure ‘work-seekers’ category.

12.3.5 Pregnant workers

In St Prix (Case C–507/12) [2015] 1 CMLR 5, the ECJ held that a woman who was obliged to (temporarily) leave employment in order to take maternity leave was entitled to retain her ‘worker’ status, and correspondingly her right of residence. The Court held that, where ‘the physical constraints of the late stages of pregnancy and the immediate aftermath of childbirth’ meant that a woman had to give up work during that period, this did not ‘deprive her of the status of “worker” within the meaning Article 45’. The Court added:

In determining what is ‘reasonable’, the national court should take account of all the ‘specific circumstances’ of the case and the ‘applicable national rules on the duration of maternity leave’.

12.3.6 Frontier workers

It is obvious from an examination of all of the cases above that Article 45 applies when a person moves from his or her ‘home’ state to work (or at least look for work) in another Member State. However, Article 45 also applies to the situation where a person lives in their ‘home’ state but works in a neighbouring state, effectively commuting to work across a national border. Such a person is known as a ‘frontier worker’. Examples include:

  • Geven (Case C–213/05) [2007] ECR I–6347 – Dutch national living in the Netherlands, commuted to work in Germany;
  • Erny (Case C–172/11) [2012] 3 CMLR 31 – French national living in France, commuted to work for the Daimler car company in Germany;
  • Giersch & Others (Case C–20/12) [2014] 1 CMLR 2 – Belgian nationals living in Belgium, and a German national resident in Germany, commuted to work in Luxembourg;
  • S & G (Case C–457/12) [2014] 3 CMLR 18; [2014] QB 1207 – Dutch nationals living in the Netherlands, commuted to work in Belgium;
  • Depesme & Others (Cases C–401–403/15) (unreported) – French nationals living in France, and a Belgian national living in Belgium, commuted to work in Luxembourg;
  • Eschenbrenner (Case C–496/15) [2017] 3 CMLR 12 – French national living in France, commuted to work in Germany;
  • Case C–802/18 F.V. & G.W. (2020) (unreported) – French national living in France, commuted to work in Luxembourg.

The ECJ has also been asked whether Article 45 applies in a less obvious ‘frontier’ situation, namely when a person has found employment in his or her ‘home’ state, but then moves their residence to another state, and commutes from there to their ‘home’ state to work. The answer is ‘yes’. The cases are:

  • Hartmann (Case C–212/05) [2007] ECR I–6303 – a German national, employed by the Post Office in Germany, moved to Austria to live but continued working in Germany;
  • Hendrix (Case C–287/05) [2007] ECR I–6909 – a Dutch national, employed at a DIY store in the Netherlands, moved to Belgium to live but continued working in the Netherlands.

In both cases, the ECJ confirmed that Article 45 applied on these facts. In Hartmann, confirmed in Hendrix, the Court stated:

The case of Hartmann will be examined in more detail below, in the section dealing with ‘social and tax advantages’. A more recent example of this phenomenon was seen in P.F. (Case C–830/18) [2021] CEC 148, in which a German national lived in France but commuted to work in his ‘home’ country. This case is also examined in more detail in the ‘social and tax advantages’ section.

12.4 Equality in social and welfare provisions and Regulation 492/2011

An important piece of secondary legislation, Regulation 1612/68, was passed in order to elaborate the meaning and scope of Article 45(2) and (3). In 2011, the Regulation was repealed and replaced by Regulation 492/2011, although the key provisions in both pieces of legislation are identical. (In most cases, even the article numbers are unchanged, although Article 12 of Regulation 1612/68 is now Article 10 of Regulation 492/2011.) This means that any ECJ case law under the 1968 Regulation remains applicable under the 2011 Regulation.

12.4.1 Eligibility for employment: Articles 3 and 4

Linguistic knowledge

The proviso to Article 3(1) was invoked in the context of schoolteachers in Groener v Minister for Education (Case 379/87) [1989] ECR 3967:

The Court stated:

Article 4(1) was infringed in Commission v France (Case 167/73) [1974] ECR 359, where French legislation, the Code du Travail Maritime 1926, imposed a ratio of 3:1 in favour of Frenchmen for the crew of French merchant ships. It was also used in Bosman (Case C–415/93) [1995] ECR I–4921, to challenge what was known as the’3+2’ rule. This case is examined in the section on sport, below.

12.4.2 Article 7(1) – prohibition of discrimination in employment

Discrimination comes in two forms: direct and indirect. Both are prohibited, although either form of discrimination could be exempted in two circumstances:

  • by Article 45(3), which deals with derogations justified on grounds of ‘public policy, public security or public health’ (discussed in the previous chapter);
  • by Article 45(4), which deals with employment in the ‘public service’ (discussed below).

Apart from that, however, direct discrimination cannot be justified. Indirect discrimination could be justified on the basis that it was necessary to satisfy some overriding national interest (this is discussed in section 12.7.

Direct discrimination

National legislation which, on its face, discriminates against workers because of their nationality, is clearly contrary to Article 45(2). Some examples are to be found in section 12.6 on ‘public service’ employment.

Indirect discrimination

National rules which are not directly discriminatory (in that they appear on their face to apply regardless of nationality) may nevertheless be in breach of Article 45(2), if the effect of those rules is to discriminate either in favour of that state’s nationals or against nationals of other states. O’Flynn (Case C–237/94) [1996] ECR I–2617 provides an example.

In O’Flynn, the ECJ explained the sort of national rules which may be found to be indirectly discriminatory:

Indirect discrimination was found in the following cases:

  • Sotgiu v Deutsche Bundespost (Case 152/73) [1974] ECR 153 – the German Post Office paid a separation allowance to all employees forced by work to live away from their families: 10 DM per day was paid to workers whose family home was in Germany; 7.5 DM per day was paid to all other employees. The ECJ found that German nationals were far more likely to qualify for the higher allowance than migrant workers.
  • Alluè & Coonan (Case 33/88) [1989] ECR 1591 – Italian legislation provided that foreign-language assistants could only be employed on year-long contracts. No such limits were imposed on other university workers. Evidence showed that 75% of foreign-language assistants in Italian universities were from outside Italy.
  • Scholz (Case C–419/92) [1994] ECR I–505 – the recruitment procedure adopted by the University of Cagliari in Italy took account of each candidate’s previous public sector employment, but only if it had been in Italy. The ECJ found that most Italians would satisfy this criterion but only a minority of migrant workers.
  • Schöning-Kougebetopoulou (Case C–15/96) [1998] ECR I–47 – German legislation conferred automatic promotion on public – sector workers (such as health care workers) after eight years’ employment – but only if that employment had been in Germany. The ECJ found that most Germans would satisfy this criterion but only a minority of migrant workers.
  • Köbler v Austria (Case C–224/01) [2003] ECR I–10239 – Austrian legislation conferred a special length-of-service increment on professors who had completed 15 years’ service in Austrian universities. The ECJ found that most Austrian professors would satisfy this criterion but only a minority of professors from other Member States.

Note: the German deutschmark (DM) ceased to be legal tender when the single European currency was introduced in 2002.

12.4.3 Article 7(2) – ‘social and tax advantages’

Article 7(2) provides that the worker ‘shall enjoy the same social and tax advantages as national workers’. Article 7(2) has been defined very widely, as the case of Fiorini v SNCF (Case 32/75) [1975] ECR 1085 demonstrates.

The Court stated:

In Even (Case 207/78) [1979] ECR 2019 the ECJ defined ‘social advantages’ as follows:

This definition has been employed in many subsequent cases, mostly involving entitlement to state benefits. However, such is the width of the interpretation given to Article 7(2) that it has extended into a number of unexpected areas:

  • the right to have a trial conducted in a particular language (Mutsch (Case 137/84) [1985] ECR 2681);
  • the right to reside with an unmarried companion (Reed (Case 59/85) [1986] ECR 1283);
  • the right to funding and maintenance to pursue full-time education (Lair (Case 39/86) [1988] ECR 3161).

Funding and maintenance for education

Lair established the proposition that Article 7(2) could be relied upon by a migrant worker to claim funding and maintenance to pursue full-time education. The rule established in Lair is controversial, as it generally entails the worker giving up their status as worker in order to pursue the course of full-time education. Hence, the ECJ has imposed limitations.

The Court stated:

Thus, in order for a migrant worker to be entitled for funding and/or maintenance grants to pursue education, there must be ‘some continuity’ unless the worker became ‘involuntarily’ unemployed.

Continuity

In the following cases the element of continuity was satisfied:

  • Matteucci (Case 235/87) [1988] ECR 5589 – period of employment as a teacher of eurythmics; followed by a singing and voice-training course;
  • Bernini (1992) – period of employment in design and planning department of a furniture factory; followed by architectural studies degree.

In Lair, the ECJ referred to ‘the previous occupational activity’, i.e. the singular, suggesting that the migrant worker’s most recent employment had to be connected to the course of study. However, the ECJ has relaxed this requirement. In Raulin (Case C–357/89) [1992] ECR I–1027, the ECJ stated that it was:

for the national court to assess whether all the occupational activities previously exercised in the host Member State, regardless of whether or not they were interrupted by periods of training or retraining, bear a relationship to the studies in question.

Involuntary unemployment

This point was addressed by the ECJ in Ninni-Orasche (2004), where an Italian worker had applied for a maintenance grant for a higher education course in languages at an Austrian university after losing her job as a waitress. There was clearly no ‘continuity’, so her claim depended entirely on her being ‘involuntarily’ unemployed. Her waitressing job had been on a fixed-term basis, and she had known this from the outset. It was argued that, because she knew before she took the job that she would subsequently lose it, her unemployment was in a sense chosen and thus ‘voluntary’. The ECJ rejected this:

Studying abroad

Assuming that the above conditions are satisfied, however, there are no other restrictions. Article 7(2) may even be relied upon to claim funding for migrant workers to study abroad. This point was made in Matteucci (1988), where the ECJ stated:

This principle applies even if it means the migrant worker returning to their home state! This point was established in Di Leo (1990), involving (what is now) Article 10 of Regulation 492/2011. The daughter of an Italian national working in Germany wanted to study medicine at the University of Sienna in Italy. German legislation made provision for grants for Germans who wished to study abroad. This was extended to the children of non-Germans working in Germany, provided that the child did not return to their home state. The ECJ held that this proviso was contrary to Article 10. Di Leo (1990) was then applied to Article 7(2) in Bernini (1992):

The Court stated:

The continuing relevance of the Lair / Bernini case law is demonstrated in these two cases.

  • Commission v Netherlands (Case C–542/09) [2012] 3 CMLR 27 – a rule of Dutch law provided for funding for higher education pursued outside the Netherlands, but only for those people who had resided lawfully in the Netherlands during at least three out of the last six years. The ECJ confirmed that ‘assistance granted for maintenance and education in order to pursue university studies … constitutes a social advantage’, following Lair and Bernini. The rule was ultimately held to be contrary to Article 7(2) – it indirectly discriminated against foreign nationals working in the Netherlands compared to Dutch nationals.
  • L.N. (Case C–46/12) [2013] 2 CMLR 37 – an EU citizen (nationality unknown), entered Denmark in June 2009 and started full-time work within a week. In August, he applied for funding in order to attend Copenhagen Business School, starting in September. However, the Danish authorities refused on the basis that, back in March, he had already applied for a place at the School and therefore his principal objective in coming to Denmark was to pursue a course of study. He challenged that refusal. The ECJ held that it was for the national court to determine whether or not he had worker status, but that if he had, he would be entitled to the funding under Article 7(2), again following Lair and Bernini.

Summary of the ‘funding and maintenance’ cases

Case Home state Host state Course
Lair (1988) France Germany Languages and literature; University of Hanover, Germany
Matteucci (1988) Italy Belgium Singing and voice-training course in Germany
Raulin (1992) France Netherlands Visual arts course at the Gerrit Rietveld Academy, Amsterdam
Bernini (1992) Italy Netherlands Architectural studies at the University of Naples, Italy
Ninni-Orasche (2004) Italy Austria Languages and literature, specialising in Italian and French; University of Klagenfurt, Austria
L.N. (2013) Unknown Denmark Copenhagen Business School

All of this case law is still relevant today, despite the rapidly expanding jurisprudence on students’ rights as Citizens to claim funding for education (discussed in Chapter 11). This is because of Article 24(2) of Directive 2004/38, which provides as follows:

You may also recall from reading Chapter 11 that ‘the right of permanent residence’ is only conferred on those who have resided continuously in the host state for five years (Article 16, Directive 2004/38). In Förster (Case C–158/07) [2008] ECR I–8507 (discussed in Chapter 11), Article 24(2) was explicitly referred to by the ECJ to support its conclusion that the five-year continuous residency period imposed under Dutch law for students to become eligible for a maintenance grant was legitimate and proportionate.

The decision of the ECJ in Förster is criticised here for failing to acknowledge that the principles in Lair, Matteucci and Bernini applied to Ms Förster. She had actually worked in the Netherlands as a primary school teacher (and later in a ‘special school providing secondary education to pupils with behavioural and/or psychiatric problems’) before embarking on her degree course in educational theory. It is fairly clear that there was a very close connection between her work and her course of study, bringing her situation into line with cases such as Lair. Ms Förster should therefore have been able to rely upon Article 7(2) of Regulation 492/2011 in order to claim her full grant. This argument was actually put to the Court by the European Commission and A-G Mazák, but the Court completely ignored it.

Unsuccessful claims

Unsuccessful claims based on Article 7(2) are rare. In Leclere and Deaconescu (2001), the ECJ held that Article 7(2) could not be invoked to challenge the refusal of a benefit where the claimant was no longer working or even resident in the country in which they had been employed. More recently, in Baldinger (Case C–386/02) [2004] ECR I–8411, the ECJ dismissed a claim based on Article 7(2) for entitlement to war compensation. This was a benefit introduced in 2000 and payable to Austrian nationals who had been taken as prisoners of war during or immediately after the Second World War. The ECJ held that this benefit could not be classed as a ‘social and tax advantage’:

Family members’ entitlement to Article 7(2) protection

This entitlement to ‘social and tax advantages’ is afforded to workers’ families too. The ECJ has pointed out that to allow Member States to discriminate against migrant workers’ families in this respect could inhibit the free movement of those workers. In the earliest case on this point, Fiorini v SNCF (1975), the ECJ allowed the Italian widow of an Italian worker in France to claim reduced rail travel, but on the basis that he would have been entitled to it had he been alive. Subsequently, in Inzirillo (Case 63/76) [1976] ECR 2057 and Castelli (Case 261/83) [1984] ECR 3199 the ECJ held that migrant workers could make claims for benefits on behalf of their families. Finally, in Bernini (1992), the ECJ went further and held that the family members themselves could make a claim under Article 7(2). The ECJ held:

The Bernini principle, that a worker’s family members can invoke Article 7(2) to claim social advantages, specifically funding for education, has been confirmed in several cases subsequently. It applies to cases in which the worker’s child claims funding from their working parent’s ‘host’ state in order to study in the child’s ‘home’ state. It has also been extended to cover cases in which the worker’s stepchild claims funding.

  • Meeusen (Case C–337/97) [1999] ECR I–3289 – the daughter of a Belgian national working in the Netherlands was able to claim funding under Article 7(2) from the Dutch authorities to study chemistry at a college in Belgium;
  • Giersch & Others (2014) – the sons and daughters of various Belgian and German nationals working in Luxembourg were able to claim funding under Article 7(2) from the Luxembourg authorities to study in Belgium and the UK;
  • Depesme & Others (Cases C–401–403/15) (unreported) – the stepchildren of two French nationals (who lived in France), and one Belgian national (who lived in Belgium) but who worked in Luxembourg, were able to claim funding from the Luxembourg authorities to study at universities in France or Belgium, respectively;
  • P.F. (Case C–830/18) [2021] CEC 148 – the son of a German national (who lived in France) but worked in Germany was able to recover the costs of school transport from the German authorities.

In Giersch & Others, the ECJ explained the situation as follows:

Finally, the nationality of the worker’s family member is irrelevant in this context (Deak (Case 94/84) [1985] ECR 1873).

Summary of cases involving Article 7(2) and family members

Case Home state(s) Host state(s) Family member(s)
Fiorini (1975) Italy France Wife and children
Inzirillo (1976) Italy Belgium Son
Castelli(1984) Italy Belgium Mother
Deak (1985) Italy Belgium Son
Bernini (1992) Italy Netherlands Daughter
Schmid (1993) Germany Belgium Daughter
Meeusen(1999) Belgium Netherlands Daughter
Giersch & Others (2014) Belgium and Germany Luxembourg Sons and daughters
Depesme & Others (Cases C-401-403/15) Belgium and France Luxembourg Step-children
P.F. (Case C-830/18) Germany Germany Son

The ECJ’s flexible approach as to whether it is the worker or a family member who claims the ‘social advantage’ can be seen in Hartmann (Case C–212/05) [2007] ECR I–6303, a case involving a claim for child-raising allowance, a benefit payable under German social security legislation. The Court confirmed that this constituted a ‘social advantage’ and, moreover, that it was immaterial which parent claimed the allowance. The Court stated (emphasis added):

12.4.4 Article 10 – access for worker’s children to education

Article 10 has been extended beyond conditions of entry to ‘general measures to facilitate attendance’, including funding in the shape of grants and loans to undertake such courses. In Casagrande (Case 9/74) [1974] ECR 773, the ECJ stated:

Age limits

For the purposes of Article 10, there is no age restriction as far as the child is concerned. This point was made in Gaal (Case C–7/94) [1995] ECR I–1031, where the ‘child’ in question was a 22-year-old biology student.

The Court stated:

In Case C–529/11 Alarape & Tijani (Case C–529/11) [2013] 3 CMLR 38; [2013] 1 WLR 2883, the facts of which were discussed in Chapter 11, a 24-year-old PhD student at Edinburgh University was nevertheless held to be the ‘child’ of a worker. Article 10 may, therefore, encompass ‘children’ aged over 21 and no longer dependent. Otherwise, students would be rendered ineligible for state financial assistance as soon as they reached 21 and were financially independent of their parents.

Step-children

Article 10 refers to the ‘children of a national of a Member State’. However, the Court has held that a worker’s step-children are included, and that these children can be of any nationality. This has been seen in two cases:

  • Baumbast (Case C–413/99) [2002] ECR I–7091 – German worker with Colombian step-daughter living in the UK: Article 10 applied.
  • Alarape & Tijani (2013) – French worker with Nigerian step-son living in the UK: Article 10 applied.

Death, retirement, etc. of the ‘worker’

In order for Article 10 to apply, it is essential that one of the child’s parents is, or has been, working in the state providing the education (Brown v Secretary of State for Scotland (Case 197/86) [1988] ECR 3205; Humbel (Case 263/86) [1988] ECR 5365). However, because Article 10 refers to the children of a worker who is ‘or has been’ employed, provided that the working parent did work in the state at some point, it is immaterial that he has since retired, died or moved on to work in another country, or become unemployed, as these cases illustrate:

  • Michel S (Case 76/72) [1973] ECR 457 – Italian national worked in Belgium until his death. Son entitled to continue to rely on Article 10.
  • Casagrande (1974) – Italian national worked in Germany until his death. Son entitled to continue to rely on Article 10.
  • Gaal (1995) – Belgian national worked in Germany until his death. Son entitled to continue to rely on Article 10.
  • Baumbast (2002) – German national worked in the UK until economic circumstances forced him to leave the UK and work in China and Lesotho in southern Africa. Daughter and step-daughter entitled to continue to rely on Article 10.
  • Ibrahim (Case C–310/08) [2010] ECR I–1065 – Danish national worked in the UK, followed by a period of unemployment, and then left the country. Children entitled to continue to rely on Article 10.
  • Ahmed (Case C–115/15) [2017] 1 CMLR 12; [2017] QB 109 – German national worked in the UK, and then left the country. Daughters entitled to continue to rely on Article 10.
  • J.D. (Case C–181/19) (2020) (unreported) – Polish national worked in Germany but was then unemployed. Daughters entitled to continue to rely on Article 10.

A similar situation exists if the child’s parents divorce and the child lives with the non-working parent (R v Home Secretary (Case C–413/99) [2002] ECR I–7091); Teixeira (Case C–480/08) [2010] ECR I–1107). What happens if the worker leaves the host state and the child accompanies him but then decides to return in order to complete their education? The ECJ has held that in such circumstances the child may be covered by Article 10, in the interests of education continuity. This occurred in Moritz (Case 389/87) [1989] ECR 723, where M, a German national, had gone to live with his father in the Netherlands when the latter worked there. Subsequently the father returned to Germany and the son again accompanied him. However, the son later applied to return to the Netherlands to complete his education. The ECJ held that M had not forfeited his rights to complete his education in the Netherlands.

Right to education abroad

Article 10 is available to secure funding even if the course is abroad, provided that such right is available to nationals of the host state. In Di Leo v Land Berlin (Case C–308/89) [1990] ECR I–4185, the ECJ stated:

‘Vocational training’

In Gravier v Liège (Case 293/83) [1985] ECR 593, the ECJ was asked to define ‘vocational training’ and duly provided a very generous interpretation of the phrase.

The Court stated:

In Blaizot v University of Liège (Case 24/86) [1988] ECR 379, the ECJ went further and held that university education could constitute ‘vocational training’ ‘where the student needs the knowledge so acquired for the pursuit of a profession, trade or employment, even if no legislative or administrative provisions make the acquisition of such knowledge a prerequisite for that purpose’.

12.5 Purely internal situations

If all the facts of a particular case occur within the territory of a single EU Member State, then EU law has no role to play. This point was first made in the context of Article 45, although it has since been confirmed and applied to Article 49 (freedom of establishment) and Article 56 (freedom to provide services) as well (these freedoms are discussed in the next chapter).

12.5.1 Workers and the ‘purely internal’ rule

This principle was applied in the following cases:

  • Iorio (Case 298/84) [1986] ECR 247 – Paolo Iorio, an Italian national, was fined for not having the proper ticket on the Rome-Palermo train. He appealed, claiming that the Italian legislation imposing the fine infringed Article 45 on the basis that it inhibited free movement, but he was unsuccessful. There was no cross-border element to the case.
  • Steen (Case C–332/90) [1992] ECR I–341 – Volker Steen, a German national, had worked for the German Post Office for 12 years. In 1985, he applied for a promotion, but withdrew from it when he discovered that this would mean classifying him as a ‘civil servant’. He complained that this classification constituted a breach of Article 45, because under German law only German nationals could be appointed as civil servants. The ECJ again held that Article 45 did not apply.

12.5.2 Workers’ family and the ‘purely internal’ rule

As with the worker, if the family member has not sought to exercise their free movement rights, EU law has no role to play. This principle has been seen in several cases:

  • Morson and Jhanjan (Case 35/82) [1982] ECR 3723 – Surinamese nationals and parents of Dutch nationals in the Netherlands could not rely upon EU law because there was only one Member State involved;
  • Dzodzi v Belgium (Case 297/88) [1990] ECR I–3763 – Togolese national married to Belgian national working in Belgium could not enforce EU law (only one Member State involved);
  • Uecker and Jacquet (Cases C–64 and 65/96) [1997] ECR I–3171 – Norwegian and Russian nationals married to German nationals working in Germany could not enforce EU law (only one Member State involved);
  • Mayeur (Case C–229/07) [2008] ECR I–8 – Peruvian national married to French national working in France could not enforce EU law (only one Member State involved).

In Carpenter (Case C–60/00) [2002] ECR I–6279, however, the ECJ ruled that EU law applied to the case of a Filipino woman married to a UK national and living and working in the UK. Although this case looks superficially identical to the cases of Dzodzi, Uecker and Jacquet above, the ECJ distinguished them. The reason was that, although the UK national lived and worked in the UK, his business involved providing services to customers in other EU Member States. The ECJ therefore held that a cross-border element had been satisfied and hence his wife could invoke EU law.

Carpenter was followed in the context of Article 45 in S & G (Case C–457/12) [2014] 3 CMLR 18; [2014] QB 1207. The case involved two Dutch nationals (D1 and D2) who lived in the Netherlands but who worked in Belgium (i.e. they were frontier workers). Whilst working in Belgium, their children were looked after by D1’s Ukrainian mother-in-law (S) and D2’s Peruvian wife (G), respectively. S and G claimed rights of residence in the Netherlands, under Directive 2004/38, as D1’s dependent relative in the ascending line and D2’s spouse, respectively, but these claims were rejected by the Dutch authorities. When S and G challenged this, the ECJ held that Directive 2004/38 was inapplicable (that legislation only conferred rights on family members of EU citizens who had taken up residence in a Member State other than their home state) but Article 45 TFEU was applicable, by analogy with Carpenter, on the basis that the denial of residence status to S and G could – potentially – interfere with D1 and D2’s rights to work in Belgium. Ultimately, it was a question for the national court to determine whether the grant of residence status to S and G was necessary in order to ‘guarantee the effective exercise’ of D1 and D2’s Article 45 rights.

12.5.3 Exceptions to the ‘purely internal’ rule

There are two significant exceptions to the ‘purely internal’ rule.

First exception: returnees

First, it does not apply to ‘returnees’, that is, persons who have left their home state to exercise free movement rights (whether under Article 45, Article 49 or Article 56, or some other provision) and have then returned to their home state. Such persons may invoke EU law (Singh (Case C–370/90) [1992] ECR I–4265). Perhaps the most frequently occurring situation where EU law may be brought into play, despite a prima facie ‘purely internal’ situation, is if the person involved had exercised freedom of movement rights previously in order to obtain a qualification abroad before returning to work in their home state. In Kraus (Case C–19/92) [1993] ECR I–1663 the Court stated:

This exception to the ‘purely internal’ rule can be seen in the following cases:

  • Knoors (Case 115/78) [1979] ECR 399 – Dutch national permitted to rely on qualification obtained in Belgium while employed in the Netherlands;
  • Bouchoucha (Case C–61/89) [1990] ECR I–3551 – French national permitted to rely on qualification obtained in the UK while employed in France;
  • Kraus (1993) – German national permitted to rely on qualification obtained in the UK while employed in Germany;
  • Fernández de Bobadilla (Case C–234/97) [1999] ECR I–4773 – Spanish national permitted to rely on qualification obtained in the UK while employed in Spain;
  • Dreessen (Case C–31/00) [2002] ECR I–663 – Belgian national studied civil engineering in Germany (at Aachen State Civil Engineering College) before returning to Belgium;
  • Brouillard (Case C–298/14) [2016] 2 CMLR 7 – Belgian national permitted to rely on his master’s degree in Law, Economics & Management, awarded by the University of Poitiers in France, in order to apply for work in Belgium;
  • B.B. (Case C–166/20) (2021) (unreported) – Lithuanian national permitted to rely on her master’s degree in pharmacy obtained in the UK on her return to Lithuania;
  • A (Case C–634/20) [2022] 3 CMLR 14 – Finnish national permitted to rely on her bachelor of medicine and bachelor of surgery from the University of Edinburgh, after her return to Finland.

Second exception: the Ruiz Zambrano principle

In Ruiz Zambrano (Case C–34/09) [2011] ECR I–1177, [2012] QB 265, the ECJ created a second exception to the ‘purely internal’ rule when it decided that EU law did apply, even though there was only one Member State (Belgium) involved. The case involved two Colombian nationals who came to live in Belgium as asylum seekers. Subsequently, the couple had two children who acquired Belgian nationality – and therefore EU citizenship – by virtue of being born in Belgium. Eventually, the Belgian authorities rejected the parents’ asylum application and ordered that the family be deported. However, the ECJ ruled that deportation of the parents would infringe the rights of the children ‘of the genuine enjoyment of the substance of the rights conferred by virtue of their status as Citizens of the Union’ (emphasis added). Not only that, but the Colombian parents (or at least one of them) was entitled to a work permit in order to be able to support their Belgian children.

This is an interesting development but its practical implications are quite modest. Subsequent case law has demonstrated that the ‘purely internal’ rule has not been abandoned. Indeed, Ruiz Zambrano was distinguished within a matter of weeks in McCarthy (Case C–434/09) [2011] ECR I–3375. This case involved Shirley McCarthy, a British national, who had resided in the UK all of her life. However, she also had Irish nationality because her mother was Irish. Shirley’s husband, George, a Jamaican national, wished to be issued with a residence permit as the spouse of an EU citizen. Shirley therefore applied for a residence permit on the basis that she was an EU citizen resident in a state other than that of her nationality (Irish). The Home Office refused. She challenged that refusal, but the ECJ confirmed the continuing application of the ‘purely internal’ rule. The Court held that neither Article 21 TFEU nor Directive 2004/38 applied to the situation of an EU citizen who had never exercised his or her free movement right and who had always resided in the Member State of which he or she was a national, even if the citizen was also a national of another Member State. It followed that the family members of such a citizen could not invoke any rights under that legislation either. Finally, the situation in Ruiz Zambrano was distinguishable, because the national measures in that case had the effect of depriving an EU citizen of ‘the genuine enjoyment of the substance of the rights conferred by virtue of that status’, which was not the situation in the present case.

McCarthy was then followed and applied in Dereci & Others (Case C–256/11) [2011] ECR I–11315, where the ECJ explained that Ruiz Zambrano dealt with the very specific situation where the denial of residence to a citizen’s family member(s) would entail the citizen having to leave the territory of the EU entirely (and not just the host Member State). As a result, the Court held that various Nigerian, Serbian, Sri Lankan, Turkish and Yugoslavian nationals were not entitled to join their Austrian relatives in Austria unless a refusal of residence would entail the Austrian relatives being forced to leave the EU entirely. The Court added that the ‘mere fact that it might appear desirable … for economic reasons or in order to keep his family together in the territory of the Union’ was insufficient to establish that a citizen would be forced to leave the Union if his or her non-EU family members were denied a right of residence. It was for the national court to decide whether or not the Ruiz Zambrano test was met. Dereci & Others has itself been followed in several ECJ cases, as follows:

  • O, S & L (Cases C–356, 357/11) [2013] 2 WLR 1093 – two men of Algerian and Ivorian nationality were not entitled to join their Finnish relatives in Finland unless a refusal of residence would entail the Finnish relatives being forced to leave the EU entirely.
  • Ymeraga & Others (Case C–87/12) [2013] 3 CMLR 33 – four Kosovan nationals were not entitled to join their Luxembourg relative in Luxembourg unless a refusal of residence would entail the Luxembourg relative being forced to leave the EU entirely.
  • C.S. (Case C–304/14) [2017] 1 CMLR 31; [2017] QB 558 – a Moroccan national (who had been convicted of a criminal offence and was subject to a deportation order) was not entitled to remain in the UK unless her expulsion would entail her son (a British national of whom she was the primary carer) being forced to leave the EU entirely.
  • Chavez-Vilchez & Others (Case C–133/15) [2017] 3 CMLR 35 – eight women (two from Venezuela, two from Suriname and one each from Cameroon, Nicaragua, Rwanda and the former Yugoslavia) would not be entitled to remain in the Netherlands (and claim benefits) unless their removal would entail their children (all of whom were Dutch nationals and of whom the women claimed to be the primary carer) being forced to leave the EU entirely.

In Chavez-Vilchez, the Dutch authorities disputed the various mothers’ claims to be the ‘primary carer’ of their children, given that, although the children all lived with their mothers, the fathers (all Dutch nationals) had apparently ‘acknowledged’ the children’s existence. The Court held that, in such cases:

12.5.4 Circumventing the ‘purely internal’ rule

Given that some cross-border element is essential, is there anything to stop a Citizen from going to work in the neighbouring State for six months and then returning home in order to then make a claim based on EU law? This was the situation in Akrich (2003). The Court held that the motives of any EU citizen intending to work in another EU Member State were irrelevant in assessing the legal situation of the couple at the time of their return to the ‘home’ Member State. Such conduct cannot constitute an abuse even if the spouse did not have a right to remain in the ‘home’ Member State at the time when the couple installed themselves in the other Member State. There would be an abuse of EU law only if the couple’s marriage had been one of ‘convenience’, that is, if it had been entered into for cynical reasons of acquiring marital status only, as opposed to genuine feelings of love.

Directive 2004/38 now provides explicitly for situations in which free movement rights are ‘abused’, such as the alleged marriage of convenience in Akrich. Article 35 provides:

12.6 The public service exemption and Article 45(4) TFEU

Member States are permitted to exclude foreign nationals from working in the ‘public service’ by virtue of Article 45(4). Member States could potentially exploit this derogation in order to limit the freedom of migrant workers to take up work in that state. It has therefore been restrictively interpreted by the ECJ. First, it only applies to access to employment, not conditions of employment after access has been granted (Sotgiu v Deutsche Bundespost (Case 152/73) [1974] ECR 153). In Brouil-lard (2016), the Court stressed that Article 45(4) does not allow Member States to restrict access to ‘public service’ employment in a case where one of its own nationals is the job applicant.

More significantly, the ECJ has held that Article 45(4) does not apply to all employment in the public service, only ‘certain activities’ involving the exercise of official authority (Sotgiu). In Commission v Belgium (Case 149/79) [1980] ECR 3881, the ECJ laid down the following test in determining whether a worker is employed in the ‘public service’:

Further guidance was given in Lawrie-Blum (1986), where the ECJ held that the derogation only applied to those posts which required ‘a special relationship of allegiance to the State on the part of persons occupying them and reciprocity of rights and duties which form the foundation of the bond of nationality’.

These principles can be seen in the following cases, where the jobs in question were all held not to be in the ‘public service’:

  • Commission v France (Case 307/84) [1986] ECR 1725 – nurses;
  • Lawrie-Blum (1986) – school teachers;
  • Alluè & Coonan (Case 33/88) [1989] ECR 1591 – university teachers;
  • Schöning-Kougebetopoulou (Case C–15/96) [1998] ECR I–47 – doctors.

Colegio de Oficiales de la Marina Mercante Española (Case C–405/01) [2003] ECR I–10391 introduced further limitations on the scope of the Article 45(4) derogation.

The ECJ stated:

In Alevizos (Case C–392/05) [2007] ECR I–3505, the ECJ considered that a Greek national and member of the Greek Air Force might be classed as working in the public service:

12.7 Justification for non-discriminatory rules

In addition to the specific Article 45 derogations, the ECJ has created a parallel set of derogations which may be pleaded by Member States to justify restrictions on the free movement of workers. The ECJ has authorised Member States to impose restrictions on the free movement of workers, provided that the national rule in question satisfies four criteria:

  1. it is non-discriminatory;
  2. it is justified by imperative requirements in the general interest;
  3. it is suitable for the attainment of the objective it pursues;
  4. it does not go beyond what is necessary in order to attain its objective (the ‘proportionality’ doctrine).

Two of the leading cases in this area are dealt with in the next section as they involve national rules curtailing free movement in the interests of sport (see Bosman (1995) and Lehtonen (2000)). Other cases where the ECJ has considered the application of the four-part test include Alluè and Coonan (1989) and Clean Car Autoservice (Case C–350/96) [1998] ECR I–2521. The former case involved a challenge to a provision of Italian legislation, which provided that foreign-language teaching assistants in Italian universities could only be employed on year-long contracts, and for a maximum of six years. It was argued by Ms Alluè (who was Spanish) and Ms Coonan (a British national), both of whom worked as foreign-language teaching assistants at the University of Venice, that this breached Article 45 as it restricted their freedom to work in Italy for as long as they wished. The Italian government sought to justify the restriction, and the ECJ applied the four-part test:

  1. Was the rule non-discriminatory? The rule applied to all foreign-language teaching assistants, including those of Italian nationality.
  2. Was it justified by imperative requirements in the general interest? The Italian government suggested that the rule was necessary in order to ensure that foreign-language teaching assistants retained sufficient familiarity with the language that they taught.
  3. Was it suitable for the attainment of the objective which it pursues? The ECJ thought not. The Court said that the danger of foreign-language teaching assistants losing contact with their mother tongue was ‘slight, in the light of the increase in cultural exchanges and improved communications’.
  4. Did it not go beyond what is necessary in order to attain its objective (the ‘proportionality’ doctrine)? Again, the Court thought not. It was ‘open to the universities in any event to check the level of assistants’ knowledge’.

In Clean Car Autoservice (1998), a challenge was brought to a provision of Austrian legislation which required companies to appoint a manager who was resident in Austria. CCA, an Austrian company, appointed a manager, a German national, who was resident in Berlin. Vienna City Council refused to register the company until he acquired a residence in Austria. CCA challenged this and the ECJ applied the four-part test:

  1. Was the rule non-discriminatory? The rule applied to all companies operating in Austria, regardless of the nationality of their managers.
  2. Was it justified by imperative requirements in the general interest? The Austrian government argued that it could be justified on two grounds. First, to ensure that the person appointed would be in a position to act effectively as manager. Second, to ensure that he could be served with notice of any fines which may be imposed upon him, and also to ensure that any fines imposed could be enforced.
  3. Was it suitable for the attainment of the objective it pursued? The ECJ decided that the residence requirement did not serve its objective. As far as the effective management point was concerned, a person could reside in the same state but be further from the place of business than someone living just over the border in a different state.
  4. Did it not go beyond what is necessary in order to attain its objective (the ‘proportionality’ doctrine)? The ECJ held that the rule was disproportionate as other means, less restrictive of the freedom of movement of workers, were available. Regarding the argument about the imposition of fines, the Court held that fines could be served at the company’s registered office instead.

12.8 Free movement of workers and professional sport

12.8.1 Introduction

EU law does apply to sport, but it is important to bear in mind the following basic propositions:

  • Article 45, on the free movement of workers, applies to team sports such as football, rugby, basketball and hockey. Individual sportsmen and women, such as golfers and tennis players, are not protected by Article 45 because they are not employed by anyone. However, they are entitled to invoke Article 56, on the freedom to provide services. For an example, see Deliège (Cases C–51/96 and C–191/97) [2000] ECR I–2549, discussed in the next chapter, at section 13.3.3).
  • EU law only applies to any type of sport if there is some economic activity. In other words, it applies to professional or semi-professional sport, but not to purely amateur sporting activities.
  • EU law does not apply to ‘questions of purely sporting interest’, such as the number of players on a team or the actual rules of the sport itself, whether in a professional, semi-professional or amateur contest.

The first ECJ case to examine the compatibility of national laws regulating sporting activity with EU law, specifically the free movement provisions, occurred in Walrave and Koch (Case 36/74) [1974] ECR 1405. The ECJ ruled that the rule of non-discrimination on grounds of nationality found in Articles 18, 45 and 56 TFEU did not affect the composition of sports teams, in particular national teams, the formation of which was ‘a question of purely sporting interest and as such has nothing to do with economic activity’. However, this clearly left open the possibility of invoking EU law to challenge other sporting rules which did have something ‘to do with economic activity’.

The next development was Donà v Mantero (Case 13/76) [1976] ECR 1333, which involved a challenge to rules of the Italian football federation, under which only Italian nationals could play in federation games. Here the ECJ ruled that Articles 18, 45 and 56 TFEU (as the case may be) could be invoked to challenge national sporting rules in the context of professional or semi-professional sport.

12.8.2 Using EU law to challenge transfer restrictions

The leading case here is the famous ‘Bosman ruling’, Bosman v Royal Belgian Football Association and Union des Associations Européennes de Football (UEFA) (1995). In this case the ECJ held that the ‘transfer fee’ system used in professional football whenever one player (if he was out of contract) moved to another club imposed an obstacle to the free movement of workers and was prohibited by Article 45(1). The expression ‘Bosman free’ to describe the free transfer of an out-of-contract football player from one club to another has now entered the vocabulary of professional football. The ECJ rejected a number of arguments advanced by the defendants to justify retention of the transfer fee system.

Note: the Belgian franc ceased to be legal tender when the single European currency was introduced in 2002.

It is also important to note that the Bosman ruling meant the abolition of transfer fees for out-of-contract football players only. One of the first players to benefit from this ruling was Steve McManaman, who transferred from Liverpool FC to Real Madrid in the summer of 1999 on a ‘Bosman free’ transfer. Players who are transferred while still under contract are potentially subject to a transfer fee – the fee operates as compensation to the ‘selling’ club for the loss of the player. In some cases these transfer fees can be very large indeed. Recent examples include the transfers of Belgian midfielder Eden Hazard from Chelsea to Real Madrid for £89 million, English forward Jack Grealish from Aston Villa to Manchester City for £100 million, Portuguese forward João Félix from Benfica to Atlético Madrid for £113 million, and French striker Kylian Mbappé from Monaco to Paris St Germain for £163 million. All four players were under contract to the ‘selling’ club at the dates of their moves and hence a transfer fee was payable.

Bosman was followed, but distinguished, in an ECJ case involving the compatibility of ‘transfer deadlines’ with EU law. While in Bosman the ECJ had held that transfer fees for out-of-contract players imposed an unjustifiable restriction on the free movement of workers, in Lehtonen (2000), the ECJ held that ‘transfer deadlines’ imposed a justifiable restriction.

Following Bosman and the abolition of transfer fees for out-of-contract players, UEFA introduced a new system of ‘transfer windows’ to try to regulate the movement of professional footballers. Under this system, players are only free to move from one club to another during the ‘close season’, which in Europe is June to August, and for one month mid-season (January). Although transfer windows clearly do restrict ‘free’ movement, they are probably justifiable by applying the precedent in Lehtonen (2000).

The ECJ has since given judgment in another case involving football which raised similar issues to those in Bosman. In Olympique Lyonnais (Case C–325/08) [2010] ECR I–2177, the issue was whether provisions of French law, requiring young footballers to pay damages to the club which had trained them in the event that the player did not subsequently sign professionally for that club, were compatible with Article 45. The ECJ held not.

12.8.3 Using EU law to challenge other sporting rules

Bosman also involved a challenge to a rule devised by UEFA, known as the ‘3+2’ rule. The rule, which was enforced by all national football associations in Europe, applied to football clubs competing in certain competitions (national championships or UEFA-organised competitions – the UEFA Cup [now the Europa League], the Champions’ Cup [now the Champions’ League] and the European Cup Winners’ Cup [now defunct]). It meant that they could only field three non-nationals plus two ‘affiliated’ players (meaning those players who had played in the country for an uninterrupted period of five years). The ECJ held that the rule constituted a breach of (what is now) Article 4 of Regulation 492/2011, which prohibits the imposition of ‘quotas’ on the employment of foreign nationals. The net result is that the ‘3+2’ rule was disapplied as far as it concerned EU nationals, although it could still be applied to non-EU nationals. UEFA suggested several justifications for the rule, all of which were rejected. One such argument was that the rule helped to maintain a connection between clubs and local players. This was rejected, the ECJ pointing out that there was no requirement that clubs had to employ local players. Even before Bosman (1995), many clubs fielded players born hundreds of miles away.

The implications of this aspect of the Bosman ruling have been enormous, especially in the English Premier League (EPL). The amount of extra revenue that has been generated in this country through the upsurge in the game’s popularity since the EPL’s introduction, higher ticket prices linked to better facilities and all-seater stadia, bigger ground capacities and Sky sponsorship, has allowed English clubs to attract large numbers of foreign nationals. Chelsea made football history in 1999 when it fielded a team containing no English players. It is quite common now for some EPL games, say Arsenal v Liverpool, to involve mostly players from EU Member States (particularly France, Portugal and Spain) or even further afield (such as Brazil or Senegal) and perhaps only two or three British players. This would have been unthinkable before Bosman. Although Brexit has made it more difficult for football clubs in the UK to sign players from EU Member States (because work permits issued by the UK Home Office are required for all overseas players, now including those from the EU), the abolition of nationality quotas by Bosman is still highly visible in the cosmopolitan playing squads in the EPL.

This aspect of Bosman was followed in the context of another sport, handball, which is very popular in Germany, in Kolpak (2003). This case involved a Slovakian national playing in Germany who found his opportunities limited by a rule allowing clubs to field only two foreigners. It was argued that the rule was justified on the ground that it was intended ‘to safeguard training organised for the benefit of young players of German nationality and to promote the German national team’. This was rejected. The rule did not benefit young German players because, following Bosman, it had already been repealed in the context of the other EU Member States and the signatories of the European Economic Area Treaty (Iceland, Liechtenstein and Norway). The ECJ held that it could not be applied to those states with which the EU had signed Association Agreements, including Slovakia, either.

A similar outcome to Kolpak was seen in Simutenkov (Case C–265/03) [2005] ECR I–2579, a case involving the EU/Russia Partnership Agreement. Igor Simutenkov, a Russian national, was playing in the Spanish football league with Tenerife. He was issued with a player’s licence identifying him simply as a non-EU player. This meant that he could be excluded from certain games where only a limited number of non-EU players could be fielded. He contested this and the ECJ held that discrimination against Russian nationals was prohibited by the partnership agreement between the EU and Russia. The same result occurred in Kahveci (Case C–152/08) [2008] ECR I–6291, involving a Turkish national playing in the Spanish football league. He successfully invoked the Association Agreement between the EU and Turkey to challenge playing restrictions imposed on him.

Note: To answer this question you will also need to refer back to Chapter 11.

SUMMARY

  • The free movement of workers is provided for by Article 45(1) TFEU. National rules which ‘preclude’, ‘deter’ or ‘dissuade’ nationals of one state going to work in another are prohibited (Graf, Casteels v British Airways).
  • Article 45(2) TFEU prohibits discrimination in employment based on nationality. It can be invoked by workers to challenge discrimination in national legislation or in the policies of individual employers (Angonese). Article 45(2) can also be invoked by employers (Innovative Technology Center, Caves Krier Frères). The prohibition of discrimination is emphasised by Regulation 492/2011, Article 7(1).
  • Discrimination can be direct or indirect. Direct discrimination occurs when a rule of national legislation or practice clearly applies different rules depending on nationality (Wood). Indirect discrimination occurs when a national rule is superficially neutral, but in practice is easier for nationals to satisfy and/or harder for non-nationals to satisfy (Sotgiu, Allué & Coonan, Scholz, Schöning-Kougebetopoulou, Köbler v Austria).
  • Direct discrimination can only be justified using the TFEU (i.e. on grounds of public policy, security and health, under Article 45(3) (see Chapter 11), or using the public service derogation in Article 45(4) – see below). Indirect discrimination is justifiable if it pursues a legitimate objective, is capable of achieving that objective and satisfies proportionality, i.e. it does not go beyond what is necessary to achieve that objective (Alluè & Coonan, Clean Car Autoservice, Lehtonen).
  • The word ‘worker’ is not defined in any EU legislation. In Lawrie-Blum, the ECJ established the conditions for a ‘worker’: provision of services, for another person, in return for monetary reward. Subsequent cases have emphasised the need for ‘remuneration’ (Trojani).
  • Part-time and/or low-paid workers qualify for Article 45 protection regardless of the level of remuneration. The minimum threshold is the provision of ‘effective and genuine’ activities (Levin, Kempf, Kranemann).
  • The following are also classed as workers: trainees (Lawrie-Blum, Bernini); frontier workers (Hartmann, Geven, Hendrix, Giersch & Others, Eschenbrenner); the previously employed (Leclere and Deaconescu); women on maternity leave, provided that they return to work within a ‘reasonable’ time period (St Prix).
  • People genuinely seeking work who can provide evidence to that effect are entitled to remain in the ‘host’ Member State indefinitely (Antonissen, Collins, G.M.A.).
  • Workers and their family members (see Chapter 11) are provided with further rights under Regulation 492/2011. Under Article 3, Member States may not make eligibility for employment more difficult for foreign nationals. However, it is permissible, if the ‘nature of the post’ requires it, that workers have certain linguistic knowledge. This applies to teaching (Groener) and banking (Angonese).
  • Under Article 4, Member States may not discriminate against other EU nationals by imposing ‘quotas’ on the number of foreign workers (Commission v France (Merchant Seamen), Bosman).
  • Article 7(2) extends the anti-discrimination principle to ‘social and tax advantages’. ‘Social advantages’ is not restricted to work-related benefits. It covers benefits which are ‘generally granted to national workers primarily because of their objective status as workers or by virtue of the mere fact of their residence on the national territory’ (Even). Examples include rail travel packages (Fiorini v SNCF), the right to have a partner (Reed), the right to funding for full-time education (Lair, Bernini, Ninni-Orasche).
  • Workers’ family members can invoke Article 7(2), regardless of their nationality (Fiorini v SNCF, Deak, Bernini, Giersch & Others).
  • However, Article 7(2) is not without some limits (Leclere and Deaconescu, Baldinger).
  • Under Article 10, workers’ children have the ‘right of access to general educational, apprenticeship and vocational training courses under the same conditions as the nationals of that State’. This extends to funding for education (Casagrande, Di Leo, Gaal).
  • There is no age restriction; a ‘child’ can be an adult (Di Leo, Gaal, Alarape & Tijani).
  • Article 10 extends to ‘all forms of education, whether vocational or general, including University courses’ (Echternach & Moritz, Di Leo, Gaal).
  • Article 10 includes a right for a child’s ‘primary carer’ to reside in the Member State with the child, irrespective of the carer’s nationality (Baumbast & R). The primary carer acquires a right of residence ‘on the sole basis’ of Article 10; the residency conditions in Directive 2004/38 do not apply (Ibrahim, J.D.). In principle, the primary carer’s right of residence ends when the child reaches the age of majority, but it may continue if necessary to enable the child to complete their education (Teixeira).
  • Purely internal situations are not covered by Article 45 (R v Saunders).
  • Even where family members are involved, if the worker has not exercised free movement rights, Article 45 does not apply (Morson & Jhanjan).
  • The purely internal rule does not apply to ‘returnees’ – workers who have exercised their right to move to another state, before returning home (Kraus). Nor does it apply where the denial of residence to a citizen’s family member(s) would entail the citizen having to leave the territory of the EU entirely (and not just the host Member State) (Ruiz Sambrano, Dereci, Chavez-Vilchez). Working abroad for a few months before returning home is enough to bring Article 45 into operation (Akrich).
  • Employment in the ‘public service’ is exempted from the freedom to work provisions by Article 45(4). This allows Member States to restrict access to public service employment; it does not justify discrimination against non-nationals in conditions of employment (Sotgiu).
  • ‘Public service’ employment involves ‘the exercise of powers conferred by public law and responsibility for safeguarding the general interests of the state’ (Commission v Belgium) and ‘a special relationship of allegiance to the state’ (Lawrie-Blum). It does not apply if public law powers are ‘exercised only sporadically, even exceptionally’ (Colegio de Oficiales de la Marina Mercante Española).
  • The following are not employed in the ‘public service’: doctors (Schöning-Kougebetopoulou); nurses (Commission v France (Nurses)); teachers (Lawrie-Blum, Bleis); university lecturers (Alluè & Coonan).
  • Article 45(4) has been held to be applicable to the captains of merchant ships (Colegio) and to air force pilots (Alevizos). Other occupations where Article 45(4) may apply include national security services, the higher levels of the civil service and the police force.
  • Article 45 applies to sport to the extent that sport is an economic activity (Walrave and Koch, Donà v Mantero). Article 45 applies to professional and semi-professional team sports.
  • Article 45 was successfully used to challenge transfer restrictions for out-of-contract footballers and for nationality quotas in football (Bosman) and the requirement that young players pay damages to their first club if they chose not to sign professionally for them (Olympique Lyonnais). But transfer deadlines used in basketball were held to be justifiable (Lehtonen).
  • Sportsmen and women who play individual sports are not ‘workers’ but are protected by Article 56, the freedom to provide services (Deliège).

Further reading

Articles

  • Busby, N, ‘Crumbs of Comfort: Pregnancy and the Status of “Worker” under EU Law’s Free Movement of Persons’ (2015) 44 ILJ 134.
  • Currie, S, ‘Pregnancy-related Employment Breaks, the Gender Dynamics of Free Movement Law and Curtailed Citizenship’ (2016) 53 CML Rev 543.
  • Gardiner, S and Welch, R, ‘Bosman – There and Back Again: The Legitimacy of Playing Quotas under European Union Sports Policy’ (2011) 17 ELJ 828.
  • Nic Shuibhne, N, ‘Reconnecting Free Movement of Workers and Equal Treatment in an Unequal Europe’ (2018) 43 EL Rev 477.
  • O’Leary, S, ‘The Curious Case of Frontier Workers and Study Finance’ (2014) 51 CML Rev 601.
  • Pennings, F, ‘Case Note on Casteels’ (2012) 49 CML Rev 1787.
  • Ristuccia, F, ‘The Right to Social Assistance of Children in Education and Their Primary Carers’ (2021) 58 CML Rev 877.
  • Tecqmenne, M, ‘Migrant Jobseekers, Right of Residence and Access to Welfare Benefits: One Step Forward, Two Steps Backwards?’ (2021) 46 EL Rev 765.
  • Van Elsuwege, P, ‘European Union Citizenship and the Purely Internal Rule Revisited’ (2011) 7 ECL Rev 308.
  • Welsh, A, ‘A Genuine Chance of Free Movement? Clarifying the “Reasonable Period of Time” and Residence Conditions for Jobseekers in G.M.A.’ (2021) 58 CML Rev 1591.

13 Freedom of establishment and the freedom to provide and receive services under Articles 49 and 56 TFEU

DOI: 10.4324/9781003218562-13

Article 49 provides for the freedom of establishment; Article 56 provides the freedom to provide (or receive) services. Both provisions are directly effective – Reyners v Belgium (Case 2/74) [1974] ECR 631 (Article 49) and Van Binsbergen (Case 33/74) [1974] ECR 1299 (Article 56).

  • A right of establishment is the right to install one’s self in another Member State, permanently or semi-permanently, on a self-employed basis, for the purpose of performing a particular activity there. It also gives companies the right to set up a branch or a subsidiary in another Member State.
  • The right to provide services allows an individual, established in one Member State, to provide their services in another Member State, on a temporary or spasmodic basis. It also allows a company, established in one Member State, to provide their services to anyone in another Member State. If necessary, it allows them to visit the other Member State, on a temporary or spasmodic basis, in order to do so.

13.1 Freedom of establishment and Article 49 TFEU

Article 49 refers to the taking up as well as the pursuit of professional activities for individual, self-employed persons. For companies, the freedom of establishment includes the right to set up and manage ‘undertakings’, in particular companies and firms; and the setting up of agencies, branches or subsidiaries; under the same conditions laid down for nationals of the state where establishment is effected. This list is non-exhaustive, and has been expanded on. In Commission v Germany (Insurance Services) (Case 205/84) [1986] ECR 3755, the ECJ suggested that ‘establishment’ could also include the presence in a state of an office managed by a company’s own staff, or a person who is independent but authorised to act on a permanent basis for the company.

It is possible to be established in two Member States at the same time. Paris Bar Council v Klopp (Case 107/83) [1984] ECR 2971 and Gebhard v Milan Bar Council (Case C–55/94) [1995] ECR I–4165 are two similar cases in which qualified German lawyers (already established in Germany) wished to set up second sets in France and Italy, respectively. They were challenged by the Paris and Milan Bar Councils but successfully invoked Article 49 to overcome any objections that they could not operate from more than one set simultaneously. These cases will be discussed in more detail in section 13.9.

13.1.1 The scope of Article 49 TFEU

Article 49 abolishes restrictions on the freedom of nationals (whether individuals or companies) to establish themselves in another Member State. The ECJ has stated that, for persons, Article 49 relates not only to the taking up of an activity as a self-employed person, but also to the pursuit of that activity, in the widest sense. The most obvious situation where Article 49 will apply is to prohibit national laws which clearly discriminate against nationals from other Member States seeking to establish themselves, perhaps by imposing extra conditions on them or by denying them rights available to nationals. This discrimination could be:

  • Direct. For example Steinhauser v City of Biarritz (Case 197/84) [1985] ECR 1819 – French law gave specific advantages to French nationals working as self-employed artists. This was successfully challenged by a German national who had travelled to Biarritz in south-west France to ply his trade as a landscape artist. See also Thijssen (Case C–42/92) [1993] ECR I–4047 – Belgian rules restricted the post of Insurance Commissioner to Belgian nationals.
  • Indirect. For example Stöber and Pereira (Cases C–4 and 5/95) [1997] ECR I–511 – German law provided child benefits to self-employed nationals only if their children were habitually resident in Germany. This was held to discriminate indirectly against foreign nationals who had come to Germany to set up a business but whose family had decided to remain at ‘home’.

However, Article 49 is not limited to prohibiting cases of nationality discrimination. Article 49 states that ‘restrictions on the freedom of establishment of nationals of a Member State in the territory of another Member State shall be prohibited’. Thus, any restriction imposed by one Member State on the freedom of establishment – even if it applies equally to that state’s own nationals – is potentially prohibited by Article 49. This point was first made by the ECJ in Klopp (1984). It has been confirmed in several cases since, including Gebhard (1995). However, the ECJ has also acknowledged in such cases that non-discriminatory national rules may be justified (see below).

In International Transport Workers’ Federation & Finnish Seamen’s Union v Viking Line (Case C–438/05) [2007] ECR I–10779, the Viking Line case, the ECJ was asked whether Article 49 could be invoked to challenge threatened strike action by a trade union (in other words, whether Article 49 had ‘horizontal’ direct effect). The Court held that it did:

In St Prix (2015) the ECJ established the principle that a woman who ceased work owing to the ‘physical constraints in the late stages of pregnancy and the aftermath of childbirth’ retained the status of ‘worker’ under Article 45, provided that she returned to the same or another employment within a ‘reasonable period’ after the birth of her child (see Chapter 12, section 12.3.5). In Dakneviciute (Case C–544/18) [2020] 1 CMLR 22, the Court applied the same principle to ‘self-employed persons’ under Article 49. This meant that Henrika Dakneviciute, a beauty therapist from Lithuania, was able to retain her status as a self-employed person during her maternity in the UK. This was essential in order to enable her to claim social security benefits.

13.1.2 Rights of entry and residence and Directive 2004/38

Self-employed persons have the same rights as all citizens to enter another Member State and to live there, to be accompanied by their family members, and even to retire in that state after their working life is over. The detailed rules are set out in Directive 2004/38, which was discussed in Chapter 11.

13.2 The problem of qualifications

13.2.1 The law prior to 2007

Although Article 49 removes restrictions on the freedom of EU nationals to establish themselves in another EU Member State, it says nothing about the various conditions which may be laid down in other states by legislation or by rules of trade or professional bodies relating to the education and/or training required to practise. These may vary widely from state to state and, left to their own devices, Member States may have been slow to recognise other states’ qualifications. There was thus a major barrier to the free movement of the self-employed, as well as employees whose trade is subject to national regulation.

To tackle this, the EU’s legislative bodies were empowered to ‘issue directives for the mutual recognition of diplomas, certificates and other evidence of formal qualifications’ (Article 53). During the 1970s and early 1980s many Directives were passed, largely in the health services field. But progress was slow, especially in heavily regulated areas like architecture (Directive 85/433 was 17 years in the making). Because harmonising by profession was proving slow, it was decided in 1984 to abandon the ‘sectoral’ approach, and instead adopt a general approach. This was not based on harmonisation of individual professions but the mutual recognition of qualifications in all areas where a higher education diploma was required. The underlying rationale was that a professional person, fully qualified in one Member State, was likely to have much the same skill, knowledge and competence as that required of a counterpart in another Member State.

The result was Directive 89/48, often referred to as the Mutual Recognition Directive. Directive 89/48 essentially created a presumption that those in possession of higher education qualifications (referred to in the directive as ‘diplomas’) which entitled them to practise a ‘regulated profession’ in one Member State were entitled to practise that profession in any other Member State, subject to certain limited derogations. Directive 89/48 was subsequently complemented by a second Mutual Recognition Directive, Directive 92/51, which applied the same principles to all post-secondary qualifications. Neither of these directives applied to professions where specific harmonising directives applied. Finally, Directive 99/42 covered the recognition of qualifications in various commercial and industrial sectors such as agriculture, carpentry, the hotel and restaurant sector, and textiles. Until 2007, therefore, there were three separate regimes:

  • Professions where specific harmonising directives existed. This covered the professions of general practitioner (GP), dentist, nurse, veterinary surgeon, midwife, pharmacist and architect.
  • All other professions, such as accountants, lawyers and bankers, where one of the Mutual Recognition Directives applied.
  • Those commercial and industrial sectors where Directive 99/42 applied.

13.2.2 The Qualifications Directive 2005/36

Those three separate regimes have now been consolidated by Directive 2005/36 (the Qualifications Directive). All the pre-existing legislation on qualifications, including the two Mutual Recognition Directives and Directive 99/42, was repealed, although some of the case law decided under these directives may still be of relevance today. The Qualifications Directive was later amended by Directive 2013/55. All references to Directive 2005/36 in the rest of this chapter are to the amended version of it. The general purpose of Directive 2005/36 is set out in Article 1, which states:

Article 2 sets out the scope of the Directive as follows:

A ‘regulated profession’

A ‘regulated profession’ is defined in Article 3(1)(a) as a

professional activity or group of professional activities, access to which, the pursuit of which, or one of the modes of pursuit of which is subject, directly or indirectly, by virtue of legislative, regulatory or administrative provisions to the possession of specific professional qualifications; in particular, the use of a professional title limited by legislative, regulatory or administrative provisions to holders of a given professional qualification shall constitute a mode of pursuit.

It follows, therefore, that it is possible that a ‘profession’ may be regulated in some Member States but not necessarily all of them. It depends on whether access to the ‘profession’ is subject in any given state to the possession of qualifications. In Aranitis (Case C–164/94) [1996] ECR I–135, a case on Directive 89/48, for example, the applicant had higher education qualifications in geology awarded in Greece. However, when he purported to rely upon these in Germany he was told that the ‘profession’ of geology was unregulated in Germany and hence the Directive did not apply. This was confirmed in the following cases:

  • Fernández de Bobadilla (Case C–234/97) [1999] ECR I–4773, another case on Directive 89/48, involving the question whether the ‘profession’ of art restorer was regulated in Spain (the ECJ held that this was a matter for the national court to decide).
  • Gräbner (Case C–294/00) [2002] ECR I–6515, a case on Directive 92/51, involving the question whether the ‘profession’ of ‘health practitioner’ was regulated in Austria (the ECJ answered this question in the negative).
  • Nasiopoulos (Case C–575/11) [2014] 1 CMLR 7, a case on Directive 2005/36, where the Court accepted that the ‘profession’ of ‘medical masseur-hydrotherapist’ was not regulated in Greece.

In Burbaud (Case C–285/01) [2003] ECR I–8219, another case on Directive 89/48, the ECJ held that, just because an activity was carried out in the public sector, it could still be regarded as a ‘regulated profession’. The case involved a Portuguese national who wished to rely upon her hospital administrator’s qualification awarded by the University of Lisbon in order to secure a managerial position within the French national health service. The French authorities refused, insisting upon possession of specific French qualifications. The Court, however, rejected the French government’s arguments and held that the definition of ‘regulated profession’ was a matter of EU law – indeed, it had to be, otherwise national legislatures could arbitrarily and unilaterally determine what was within the scope of the EU legislation on qualifications.

Article 3(2) of the Qualifications Directive adds that a profession practised by the members of an association or organisation listed in Annex I shall be treated as a ‘regulated profession’. This includes, in the UK, bodies as diverse as the British Computer Society, the Institute of Chartered Accountants, the Royal Institution of Chartered Surveyors, the Royal Aeronautical Society and the Engineering Council.

Article 4 elaborates on the ‘effect’ of the Qualifications Directive:

‘Partial access’ is defined, in paragraph 4f, as follows:

Directive 2005/36 then divides into two broad areas:

  • the free provision of services (Title II; Articles 5–9);
  • freedom of establishment (Title III; Articles 10–52).

The Qualifications Directive is a lengthy, detailed and technical piece of legislation and the following is necessarily only an overview of the main provisions.

The free provision of services

The main substantive provision here is Article 5, which states:

Article 5(2) was discussed in T.P. (Case C–502/20) (2021), unreported. The Court held that T.P., a Belgian national, who had practised as an automotive expert in Belgium for over 25 years before moving to establish himself in Luxembourg, was able to rely on Article 5(2) in order to provide his services ‘on a temporary and occasional basis’ back “home” in Belgium.

Article 6 adds that Member States may exempt service providers from the requirements which it places on professionals established in its territory relating to authorisation by, registration with or membership of a professional organisation or body. Article 7 allows Member States to require that, where a service provider first moves from one state to another in order to provide services, he or she informs the competent authority of the host state, in writing, in advance of doing so. Article 8 provides that the competent authorities of the host state may ask the competent authorities of the service provider’s state of establishment ‘in the event of justified doubts’ to provide any information relevant to:

  • the legality of that establishment;
  • the service provider’s ‘good conduct’; and
  • the absence of any disciplinary or criminal sanctions of a professional nature.

Under Article 9, the competent authorities of the host state may also require the service provider to provide the service recipient with various information, such as:

  • the name and address of the competent supervisory authority (if any) in the state of establishment;
  • the service provider’s professional title or formal qualification and the state in which it was awarded;
  • details of any insurance cover or other means of professional liability protection.

Freedom of establishment

The Title on establishment further subdivides into three chapters:

  • general system for the recognition of evidence of training (Chapter 1; Articles 10–15);
  • recognition of professional experience (Chapter 2; Articles 16–20);
  • recognition on the basis of co-ordination of minimum training conditions (Chapter 3; Articles 21–52).

General system for the recognition of evidence of training (Chapter 1)

This Chapter applies to all ‘regulated professions’ not covered by Chapters 2 and 3 (Article 10). It is this Chapter which replaces the Mutual Recognition Directives. It divides professional competence into five levels:

  • Level 1 – an ‘attestation of competence’ issued by a competent authority in the home state on the basis of a training course not forming part of a certificate or diploma, or three years’ full-time professional experience, or general primary or secondary education, attesting that the holder has acquired ‘general knowledge’ (Article 11(a)).
  • Level 2 – a ‘certificate’ attesting to successful completion of a secondary course which is either ‘general’ or ‘technical or professional’ in character, and which in either case is supplemented by a course of study or professional training (Article 11(b)).
  • Level 3 – a ‘diploma’ certifying either (i) successful completion of training at post-secondary level of at least one year’s duration (full-time) or the equivalent part-time duration, or (ii) regulated education and training or, in the case of regulated professions, vocational training with a ‘special structure’, if such training provides a comparable professional standard and which prepares the trainee for a ‘comparable level of responsibilities and functions’ (Article 11I).
  • Level 4 – a ‘diploma’ certifying successful completion of training at post-secondary level of at least three and not more than four years’ duration (full-time), or the equivalent part-time duration, at a university or equivalent establishment, plus any additional professional training which may be required (Article 11(d)).
  • Level 5 – a ‘diploma’ certifying successful completion of a post-secondary course of at least four years’ duration (full-time), or the equivalent part-time duration, at a university or equivalent establishment, plus successful completion of any additional professional training required (Article 11(e)).

Conditions for recognition: Article 13

Article 13 is the main substantive provision of Chapter 2. It provides:

In Toki (Case C–424/09) [2011] ECR I–2587, a case on Directive 89/48, the ECJ was asked what was meant by the reference to applicants who have ‘pursued the profession’ in (what is now) Article 13(2) of Directive 2005/36. (This is crucial in cases where the applicant has moved from a state where the profession is not ‘regulated’.) The Court held that it required ‘the continuous and regular pursuit of a range of professional activities which characterise the profession concerned’. On the other hand, time spent conducting research and teaching did not count as pursuit of the profession in question.

‘Compensation measures’: Article 14

Article 14 provides for ‘compensation measures’. It states:

In Beuttenmüller (Case C–102/02) [2004] ECR I–5405, a case under Directive 89/48, the ECJ discussed the role of the ‘compensation measures’ under that Directive (which are essentially the same as those set out above). The Court stated:

In Colegio de Ingenieros de Caminos, Canales y Puertos (Case C–330/03) [2006] ECR I–801, another case under Directive 89/48, the ECJ again examined the role of the ‘compensation measures’. The Court concluded that, just because a host Member State was entitled to require an applicant from another state to satisfy one of the measures before being allowed to practise a profession in the host state, it did not follow that it was obliged to do so. Instead, the Court created the possibility of ‘partial recognition’ of an applicant’s qualifications, allowing him or her to practise in the areas in which they were actually qualified.

The Court stated:

However, the Court added that the possibility of ‘partial taking-up’ would only be available in cases where there were such significant differences between the home and host Member States that compensation measures would be tantamount to ‘the full programme of education and training’ in the latter state. Mr Imo’s case was a perfect example: the Spanish Supreme Court had already found that there were significant differences between the profession of hydraulic engineer in Italy and that of civil engineer in Spain, such that requiring Mr Imo to undergo compensation measures in Spain would, in effect, force him to ‘complete a fresh, complete programme of education and training’. That would be disproportionate, and hence partial taking-up was a more appropriate solution. By way of contrast, where a profession in two Member States was sufficiently similar (albeit not identical), then compensation measures would be more appropriate and ‘partial taking-up’ would not be available.

In Colegio, the ECJ also acknowledged that ‘partial taking-up’ of a profession created a risk of ‘fragmenting the professions regulated in a Member State into various activities [leading] to a risk of confusion in the minds of the recipients of services, who might well be misled as to the scope of those qualifications’. However, the Court held that a blanket ban on ‘partial taking-up’ would be disproportionate. There were other, less-restrictive methods of protecting consumers from the ‘risk of confusion’.

Colegio was followed and applied in Nasiopoulos (2014), involving a Greek national who qualified as a ‘medical masseur-hydrotherapist’ in Germany and who wished to practise as a physiotherapist in Greece. The ECJ held that it was for the Greek courts to examine the differences between the two professions (i.e. physiotherapy in Greece and medical masseur-hydrotherapy in Germany). If they were ‘so great’ that requiring Mr Nasiopoulos to undergo compensation measures would effectively require him to ‘follow a full programme of education and training’ in order to pursue the profession of physiotherapist in Greece, then ‘partial taking-up’ should be made available to him; otherwise, he would have to accept the compensation measures.

Generally speaking, where one of the ‘compensation measures’ is required, then the individual applicant can choose either to undergo an ‘adaptation period’ or take an ‘aptitude test’ (full definitions of these expressions are given below). By way of derogation from this, Article 14(3) provides that host states can specify one or the other for those professions requiring ‘precise knowledge of national law’. This most obviously includes judges and lawyers – but does it extend to other professions as well? In Price (Case C–149/05) [2006] ECR I–7691, a case under Directive 89/48 involving an auctioneer, the ECJ considered what was meant by the phrase ‘precise knowledge of national law’. The Court stated:

In Van Leuken (Case C–197/06) [2008] ECR I–2627, another case under Directive 89/48, the ECJ held that the ‘profession’ of estate agent in Belgium was not one ‘whose practice requires precise knowledge of national law’. The Court pointed out:

it is sufficient to be the holder of a Belgian diploma in civil, agricultural, technical or industrial engineering in order to become a member of the profession of estate agent in Belgium and the education and training leading to those diplomas does not include significant legal training.

This meant that the Belgian authorities could not demand that Willem Van Leuken – a Dutch estate agent based in the Netherlands whose activities included selling property in Belgium to Dutch clients – undertake an aptitude test in Belgian property law. More recently, in Vandorou & Others (Cases C–422, 425 and 426/09) [2010] ECR I–12411, the Court accepted that accountancy was another profession requiring ‘precise knowledge of national law’. This meant that, although Vassiliki Vandorou had qualified as a chartered accountant in the UK, she would have to pass an aptitude test (in Greek company, commercial, employment and tax law) before being allowed to practise as an accountant in Greece.

‘Adaptation period’: Article 3(1)(g)

An ‘adaptation period’ is defined in Article 3(1)(g) of the Qualifications Directive as:

the pursuit of a regulated profession in the host Member State under the responsibility of a qualified member of that profession, such period of supervised practice possibly being accompanied by further training. This period of supervised practice shall be the subject of an assessment. The detailed rules governing the adaptation period and its assessment as well as the status of a migrant under supervision shall be laid down by the competent authority in the host Member State.

‘Aptitude test’: Article 3(1)(h)

An ‘aptitude test’ is defined in Article 3(1)(h) of the Qualifications Directive as a

test of the professional knowledge, skills and competences of the applicant, carried out or recognised by the competent authorities of the host Member State with the aim of assessing the ability of the applicant to pursue a regulated profession in that Member State. In order to permit this test to be carried out, the competent authorities shall draw up a list of subjects which, on the basis of a comparison of the education and training required in the Member State and that received by the applicant, are not covered by the diploma or other evidence of formal qualifications possessed by the applicant.

Article 3(1)(h) goes on to provide that the test must

take account of the fact that the applicant is a qualified professional in the home Member State or the Member State from which the applicant comes. It shall cover subjects to be selected from those on the list, knowledge of which is essential in order to be able to pursue the profession in question in the host Member State. The test may also cover knowledge of the professional rules applicable to the activities in question in the host Member State.

Recognition of professional experience (Chapter 2)

This Chapter deals with those commercial and industrial sectors where previously Directive 99/42 applied. Article 16 now provides that if, in a Member State, access to or pursuit of one of the activities listed in Annex IV of the Qualifications Directive is contingent upon possession of ‘general, commercial or professional knowledge and aptitudes’, that state must recognise previous pursuit of the activity in another state as sufficient proof of such knowledge and aptitudes. The list of ‘activities’ in Annex IV is very lengthy but, by way of example, it includes agriculture, carpentry, construction, footwear and clothing manufacturers, and the beverage, chemicals and petroleum industries. Articles 17–19 stipulate various minimum time periods over which the activities must have been pursued in the home state.

Recognition on the basis of co-ordination of minimum training conditions (Chapter 3)

This Chapter deals with those professions which previously had their own Directive. They now have their own provisions in the Qualifications Directive (GPs are covered by Articles 24–30; nurses by Articles 31–33a; dentists by Articles 34–37; vets by Articles 38 and 39; midwives by Articles 40–43b; pharmacists by Articles 44 and 45; architects by Articles 46–49), but the essential principles are the same for each of these areas. The Qualifications Directive:

  • co-ordinates across all states the training required in order for qualification (as a nurse, dentist, architect, etc.); and
  • provides that, once a professional in one of these areas is qualified to practise in one Member State then they are automatically entitled to practise in any other state.

13.2.3 Obligation to assess equivalence

Prior to Directive 89/48, many cases arose involving professions where no separate harmonising Directive existed. The ECJ has developed a line of case law according to which it is unlawful discrimination, and a breach of Articles 49 or 56 TFEU, to refuse permission to practise to a person whose qualifications in one state have been recognised, by the competent authorities, as equivalent to those awarded in the state in which he seeks to practise. Examples of this approach are Thieffry (Case 71/76) [1977] ECR 765 and Patrick v Minister of Cultural Affairs (Case 11/77) [1977] ECR 1199.

The ECJ then developed this rule to impose an obligation on the authorities of the host state to compare the candidate’s qualifications with those awarded in the host state in order to establish whether or not they were equivalent (UNECTEF v Heylens (Case 222/86) [1987] ECR 4097). The person concerned must be given reasons why their qualifications are not deemed to be equivalent, and the decision must be subject to judicial review. The leading case in this area is Vlassopoulou (Case C–340/89) [1991] ECR I–2357.

The ECJ held that the competent authorities are obliged to make:

The ECJ laid down some guidelines as to what Member States’ authorities should do after the assessment was completed. It distinguished between three situations: full, partial and no equivalence:

  • Full equivalence: the Member State had to accept the candidate’s qualification.
  • Partial equivalence: the candidate might be required to show that he had acquired the knowledge and qualifications lacking. Any examination of documents made in order to discover which was the case had to be done in such a way as to ensure that the person’s EU legal rights were protected. Any decision taken after the examination had to be capable of judicial review. The candidate had to be able to ascertain the reasons for any adverse decision, and also told of the remedies available against such a decision.
  • No equivalence: the Member State was under no obligation to allow the candidate to practise.

These principles remain relevant today, notwithstanding the introduction of both Directive 89/48 and Directive 2005/36. The reason for this apparently incongruous situation is that there are various situations in which a person may not be able to invoke the secondary legislation on qualifications, and hence the principles established in Thieffry, Patrick and Vlassopoulou are still important. There are at least two situations where this is possible:

  • where an individual does not have a ‘diploma’;
  • where a profession is unregulated in the ‘host’ state.

The case law principles have been applied in several cases over the years (Aguirre Borrell and Others (Case C–104/91) [1992] ECR I–3003; Fernández De Bobadilla (1999); Dreessen (Case C–31/00) [2002] ECR I–663); Morgenbesser (Case C–313/01) [2003] ECR I–13467; Peśla (Case C–345/08) [2009] ECR I–11677; B.B. (Case C–166/20) (2021), unreported; A (Case C–634/20) [2022] 3 CMLR 14. In Dreessen (2002), the ECJ summarised the position as follows:

In Morgenbesser (2003), Christine Morgenbesser, a French national, had studied law in France and obtained her diploma of maîtrise en droit in 1996, signifying completion of the academic stage of legal education. She did not, however, go on to study for her certificate d’aptitude à la profession d’avocat which would have allowed her to practise as a lawyer in France. Instead, she spent several months working in a law firm in Paris. Then, in 1998, she joined a firm of lawyers (avvocati) in Genoa, Italy. In 1999, she applied for enrolment in the register of ‘praticanti’ – lawyers who have the academic qualifications but not the practical training. However, her application was refused by both the Genoa Bar Council and the National Bar Council on the basis that Italian legislation required that praticanti held a diploma awarded or confirmed by an Italian university. She challenged this.

The ECJ rejected her claim to rely upon Directive 89/48, on the basis that she did not have a ‘diploma’ as defined in that legislation. However, the Court went on to hold that, under Article 49, the Italian authorities were nevertheless obliged to undertake an examination of her French qualifications (namely, her maîtrise en droit) plus whatever practical experience she had acquired, and to assess that for equivalence to the Italian qualifications. She would then be in a position, if she wished, to undertake whatever additional education and/or practical training was required in order to become fully qualified as an avvocato in Italy.

In B.B. (Case C–166/20) (2021), a Lithuanian national had obtained a master’s degree in pharmacy in the UK in 2013, but had to return to Lithuania before completing the required postgrad professional practice to become fully qualified. This meant that she did not have the full ‘diploma’ and hence that Directive 2005/36 was not available to her. However, she could rely on Article 49 and the case law principles set out above – the Lithuanian authorities were required to take into consideration ‘all the diplomas, certificates and other evidence of formal qualifications’ and ‘relevant experience’ acquired by B.B., ‘comparing the specialised knowledge and abilities so certified, and that experience, with the knowledge and qualifications required by the national legislation’. The facts of A (Case C–634/20) (2022) were very similar – this time, a Finnish national had obtained a bachelor of medicine and bachelor of surgery from the University of Edinburgh. This gave her a restricted right to pursue the profession of doctor in the UK. She was entered as a ‘provisionally registered doctor with a licence to practise’ in the General Medical Council’s register, but at this point she returned to Finland. Not being fully qualified to practise in the UK meant that, like the Lithuanian pharmacist in B.B., she did not have the full ‘diploma’ and hence Directive 2005/36 was not available to her, either. However, she could also rely on Article 49 and the case law principles set out above to require the Finnish authorities to compare her British qualifications with those in Finland.

The principles established in Vlassopoulou were applied in a slightly different context in Vandorou & Others (Cases C–422, 425 and 426/09) [2010] ECR I–12411. Three Greek nationals, who had obtained qualifications (in accountancy and engineering) in other Member States (the UK and Germany), sought to rely on them in order to practise professionally back ‘home’ in Greece. However, because of substantial differences between the education and training undertaken abroad and that required in Greece, the Greek authorities insisted on compensation measures being undertaken. The applicants asked for practical experience which they had acquired whilst studying for their diplomas to be taken into account when determining what form the compensation measures should take. The Greek authorities refused, but the ECJ held that such experience had to be recognised, stating:

insofar as all practical experience in the pursuit of related activities can increase an applicant’s knowledge, it is incumbent on the competent national authorities to take into consideration all practical experience of use in the pursuit of the profession to which access is sought.

13.2.4 Qualifications obtained outside the EU

In Tawil-Albertini (Case C–154/93) [1994] ECR I–451, where the candidate, a French national, had a qualification in dentistry awarded in Lebanon and wanted to practise in France, the ECJ held that the Member States were not obliged to recognise qualifications obtained in a country outside the EU. This was the case even if one Member State recognises the non-EU qualification as equivalent to its own: this does not bind the other Member States. Thus, the fact that the Belgian authorities were prepared to recognise the candidate’s Lebanese qualification did not oblige the authorities in France to recognise it. However, if one Member State does recognise non-EU qualifications, and allows the person to practise there, then the other Member States are required to recognise that practical experience. This was decided in Haim (Case C–319/92) [1994] ECR I–425, where an Italian national acquired a Turkish diploma in dentistry before practising in Belgium, where his diploma was recognised, and then sought to rely upon his diploma in Germany, where it was not recognised. The ECJ held that the German authorities had to take into account Haim’s professional experience acquired in Belgium. This point was confirmed in Hocsman (Case C–238/98) [2000] ECR I–6623. Moreover, in Hocsman the ECJ seemed to say that the authorities would be obliged to recognise all qualifications awarded ‘abroad’. This would amount to an overruling of Tawil-Albertini (1994), although the ECJ did not make this explicit.

The ECJ held:

Summary of cases involving the recognition of non-EU qualifications:

Case Year Nationality of complainant Non-EU state and qualification obtained Member State where qualification recognised Member State refusing to recognise qualification
Tawil-Albertini 1994 French Lebanon – dentistry Belgium France
Haim 1994 Italian Turkey – dentistry Belgium Germany
Hocsman 2000 Spanish Argentina – medicine Spain France

These case law developments have now been recognised in legislation. Article 3(3) of the Qualifications Directive provides:

13.3 The freedom to provide services under Article 56 TFEU

Article 56 provides that ‘restrictions on freedom to provide services … shall be prohibited in respect of nationals of Member States who are established in a [Member State] other than that of the person for whom the services are provided’.

13.3.1 ‘Services’

This is defined in Article 57: ‘

Article 56 therefore has a very wide scope. It covers all commercial activities, financial and legal advice, medical treatment, even holidays. SPUC v Grogan (Case C–159/90) [1991] ECR I–4685 established that abortions provided by private clinics amounted to ‘services’ for the purpose of Article 56. In Jany and Others (Case C–268/99) [2001] ECR I–8615 the ECJ held that Article 56 could apply to prostitution!

13.3.2 ‘Freedom to provide services’

Article 56 prohibits national laws which discriminate against service providers from other Member States. This discrimination may be direct or indirect. In fact, most of the cases involve indirect discrimination. For example, in Van Binsbergen (Case 33/74) [1974] ECR 1299, Dutch rules required lawyers to be habitually resident in the Netherlands. This was held to discriminate indirectly against foreign nationals who were less likely to be living in the Netherlands. The ECJ stated:

A similar case of indirect discrimination is Coenen (Case 39/75) [1976] ECR 1547 – Dutch rules required insurance brokers to be habitually resident in the Netherlands. A more recent example is Bickel and Franz (Case C–274/96) [1998] ECR I–7637.

Article 56 may even be invoked to challenge all national legislation which is not discriminatory at all (directly or indirectly), provided that it has the potential to inhibit the freedom to provide services. This was clearly acknowledged in Säger v Dennemeyer & Co. (Case C–76/90) [1991] ECR I–4221, where the ECJ stated:

As discussed earlier in the context of Directive 2005/36, a provider of services is entitled to do so in their ‘home’ State if they have established themselves in another Member State (T.P. (Case C–502/20) (2021)). In Scorpio Konzertproduktionen (Case C–290/04) [2006] ECR I–9461, the ECJ offered the following guidance on when Article 56 applies:

The Services Directive 2006/123

In 2006, the Council adopted the Services Directive (Directive 2006/123). Although popularly known as the Services Directive it actually applies to Establishment as well. The purpose behind the Directive is to force Member States to become more proactive in identifying and removing barriers to service providers seeking either to establish themselves in another Member State or to provide services in another state. The preamble to the Directive explains this:

The key points are:

  • The Directive apples to services supplied by providers established in a Member State (Article 2(1)).
  • A number of services are excluded, including ‘non-economic services of general interest’, financial services, electronic communications services, transport, temporary work agencies, healthcare, gambling, audiovisual, private security and social services (Article 2(2)).
  • Otherwise, the Directive applies to ‘any self-employed economic activity, normally provided for remuneration’ (Article 4(1)).
  • The Directive provides that cross-border service providers must be free to do so, except where it is justifiable to restrict that freedom ‘for reasons of public policy, public security, public health or the protection of the environment’ (Article 16). These derogations must ‘respect the principles’ of ‘non-discrimination: the requirement may be neither directly nor indirectly discriminatory with regard to nationality or, in the case of legal persons, with regard to the Member State in which they are established’ and ‘proportionality: the requirement must be suitable for attaining the objective pursued, and must not go beyond what is necessary to attain that objective’ (Article 16).

13.3.3 ‘Remuneration’

Article 56 applies whenever a cross-border service is provided in return for remuneration. In the vast majority of cases this proposition will be straightforward, but in certain circumstances it has prompted questions that needed to be resolved by the ECJ. For example, in Bond van Adverteerders (Case 352/85) [1988] ECR 2085, subsequently confirmed in Deliège (Cases C–51/96 and C–191/97) [2000] ECR I–2549, the ECJ held that it is not essential that the person who receives the services be the person who provides the remuneration.

Two areas of cross-border provision of services have generated controversy, because of the question of remuneration.

Educational services

In Humbel (Case 263/86) [1988] ECR 5365 the ECJ made it very clear that while privately funded education is covered by Article 56, state-funded education is not. Humbel (1988) has been followed ever since. In Wirth (Case C–109/92) [1993] ECR I–6447, the ECJ held that courses ‘given in an establishment of higher education which is financed essentially out of public funds’ do not constitute ‘services’, while in Schwarz (Case C–76/05) [2007] ECR I–6849, the Court stated:

Article 56 is applicable where taxpayers of a given Member State send their children to a private school established in another Member State which may be regarded as providing services for remuneration, that is to say which is essentially financed by private funds.

Medical treatment

Luisi and Carbone (Cases 286/82 and 26/83) [1984] ECR 377 established that privately funded medical treatment is covered by Article 56. The principle that private medical treatment is a ‘service’ and therefore subject to Article 56 has been confirmed in several cases since, for example Watts v Bedford Primary Care Trust (Case C–372/04) [2006] ECR I–4325. The case involved Yvonne Watts, a British woman who underwent a hip replacement operation at a private hospital in France, because she would have had to wait for a year to have the operation in the UK. Regarding the question whether Article 56 applied on these facts, the ECJ stated:

13.3.4 What if the ‘service’ is illegal?

This issue arises because a service may be permitted in one Member State but prohibited (or at least restricted) in another. Article 56 applies provided that the service is lawful in some, but not necessarily all, Member States. The best example is perhaps an English case that did not even reach the ECJ. R v Human Fertilisation and Embryology Authority, ex p Blood [1997] 2 CMLR 591 involved the famous case of Diane Blood. Her husband, Stephen, suddenly contracted meningitis and lapsed into a coma (from which he never recovered). The couple were childless but they had both wanted children. Mrs Blood therefore applied to the High Court for permission to use Stephen’s sperm for artificial insemination. However, the Court refused because UK legislation, the Human Fertilisation and Embryology Act 1990, requires the written consent of the man concerned, and Stephen had not given this consent. Mrs Blood appealed and the Court of Appeal held that she had a right under Article 56 to take Stephen’s sperm for insemination treatment in Belgium, where written consent was not required. Treatment was duly carried out, successfully, in 1998 and Mrs Blood had her first child in 1999.

Society for the Protection of Unborn Children (SPUC) v Grogan (1991) raised the question of whether Article 56 could be relied upon by students in one Member State (Ireland) to disseminate information about abortion clinics in another Member State (the UK), given that, at the time, abortion was absolutely prohibited in Ireland.

13.3.5 The scope of ‘cross-border’ services

Although the provider of services will typically travel to another state to do so, it is not necessary. Article 56 also applies where the service recipient travels to a state other than their own in order to receive a service – most obviously, tourists and those seeking specialist medical treatment (see section 13.5). Indeed, it is not actually necessary for either the service provider or the recipient to leave their own state. Provided that they are established in different states, and the service is provided by one and received by the other, then Article 56 will apply. This allows service providers to challenge measures adopted by their own state where those measures prohibit or restrict their freedom to provide services to nationals established in other Member States. This is increasingly likely to become the case where instantaneous telecommunication systems – most obviously the Internet – are relied upon. In Alpine Investments (Case C–384/93) [1995] ECR I–1141, the ECJ stated that Article 56 ‘covers services which the provider offers by telephone to potential recipients established in other Member States and provides without moving from the Member State in which he is established’. In Carpenter (Case C–60/00) [2002] ECR I–6279, the facts of which were given in the previous chapter (see section 12.5.2), the ECJ confirmed Alpine Investments (1995) and stated:

Services come within the meaning of ‘services’ in [Article 56] both in so far as the provider travels for that purpose to the Member State of the recipient and in so far as he provides cross-border services without leaving the Member State in which he is established.

13.4 Distinguishing establishment and services

The difference between the two is a matter of degree – whether someone is simply providing services, or is actually established, in a Member State other than their own may simply depend on how lengthy is their stay in that other state. It is sometimes difficult to differentiate between established persons and service providers. Occasionally, the ECJ may need to decide. The leading case is Gebhard (1995), where the ECJ stated:

In Schnitzer (Case C–215/01) [2003] ECR I–14847, the ECJ offered further guidance on the difference between establishment and the provision of services. The Court stated that Article 56 may cover services ‘varying widely in nature’, including services which are provided ‘over an extended period, even over several years, where, for example, the services in question are supplied in connection with the construction of a large building’. The Court then stated:

The ECJ has held that, where an enterprise is situated in one Member State, but its activities are directed entirely or mainly towards another Member State (often in order to avoid more stringent rules and regulations applying in the latter state), then the enterprise will be deemed to be established in the latter state and not simply providing services there (TV 10 (Case C–23/93) ECR I–4795).

More recently, in Garkalns (Case C–470/11) [2012] 3 CMLR 52, the Court held that Article 49 (establishment) applies when an ‘operator offers its services on a stable and continuous basis from an establishment in the Member State of destination’; otherwise Article 56 (services) applies. Here, Article 49 applied to the proposed development of an amusement arcade in a shopping centre in Riga, the capital of Latvia, as the gambling services in question would eventually be offered ‘on a stable and continuous basis’.

Although ‘the right of establishment [and] the chapter on services [are] mutually exclusive’ (Gebhard), it is possible for the same case to involve both establishment and services. In P.I. (Case C–230/18) [2019] 3 CMLR 31, P.I., a Bulgarian national, owned and ran a massage parlour in Innsbruck, Austria. In December 2017, the local police suspected that it was a front for prostitution and that naked and erotic massages were bring provided, and ordered that it be shut. P.I. challenged that on the basis that her EU rights had been infringed. The question is, which right(s)? The Court of Justice held that, whilst establishing and operating a massage salon in another Member State fell within under Article 49, any prostitution services that were (allegedly) being carried out as well fell under Article 56. In the end, the case was dealt with under Article 49.

13.5 The freedom to receive services

Although Article 56 refers to the right to provide services, it has been extended to the right to receive them too. The first case was Luisi and Carbone (1984).

The Court stated:

Note: The French franc and the German Deutschmark ceased to be legal tender when the single European currency was introduced in 2002.

This ruling has since been followed in several cases:

  • tourists – Cowan (Case 186/87) [1989] ECR 195; Bickel and Franz (1998); Calfa (Case C–348/96) [1999] ECR I–11;
  • persons receiving private education (see above);
  • persons receiving medical treatment (see above).

A variation on this theme occurred in De Coster (Case C–17/00) [2001] ECR I–9445, which involved the imposition of a satellite dish tax by a local authority in Belgium. One of the residents challenged this, alleging that it restricted his freedom to receive services (in the form of satellite television broadcasts) from other Member States. The ECJ agreed (although it found the tax was at least potentially justifiable because it helped to control the proliferation of satellite dishes, which was beneficial to the environment).

13.6 The ‘official authority’ derogation in Article 51 TFEU

Article 51 provides that ‘activities … that are connected, even occasionally, with the exercise of official authority’ are not covered by the Treaty rules on establishment and services. The derogation has, however, been given a very narrow scope. The ECJ has pointed out that Article 51 applies only to activities connected with the exercise of official authority; not professions or occupations as a whole. In Reyners v Belgium (Case 2/74) [1974] ECR 631, involving Belgian law restricting the profession of avocat (similar to barrister) to Belgian nationals, the ECJ stated:

A similar outcome was seen in Peñarroja (Cases C–372 and 373/09) [2011] ECR I–1785, where the Court held that Article 51 did not apply to the activities of court translators. However, in both cases, the Court implied that, although the activities of lawyers and court translators did not involve the ‘exercise of official authority’, the activities of judges do. In Reyners, for example, the Court stated that ‘the exercise of these activities leaves the discretion of judicial authority and the free exercise of judicial power intact’.

The ECJ continues to insist that only activities involving a ‘direct and specific’ connection with official authority are covered by Article 51. In six cases decided on the same day in May 2011 the ECJ agreed with the European Commission that six Member States – Austria, Belgium, France, Germany, Greece and Luxembourg – had breached Article 49 by requiring that notaries hold the nationality of the state concerned. Notaries in civil law countries (as opposed to common law jurisdictions) may be defined as legally qualified, public officials appointed by private parties to authenticate legal documents such as contracts, conveyances, deeds, gifts, powers of attorney and so on. In all six cases, the defendant state argued that the profession was exempted by Article 51, but the Court pointed out that the typical activity of a notary (authenticating documents), could not be said to involve a ‘direct and specific’ connection with the ‘exercise of official authority’ (see, for example Commission v Germany (Case C–54/08) [2011] ECR I–4355).

In Thijssen (Case C–42/92) [1993] ECR I–4047, the ECJ imposed another limitation on the scope of Article 51. The Court had been asked whether the post of Insurance Commissioner in Belgium was one which involved the ‘exercise of official authority’. The Court answered ‘no’, on the basis that the post involved only ‘auxiliary and preparatory functions’. It was the Belgian Insurance Inspectorate itself which had decision-making powers and therefore exercised official authority, rather than its individual employees. This limitation has been applied in other cases involving inspectors supervised by public authorities. In Commission v Austria (Case C–393/05) [2007] ECR I–10195 and Commission v Portugal (Case C–438/08) [2009] ECR I–10219, involving organic farm inspectors and motor vehicle inspectors, respectively, the Court held that Article 51 did not apply because the inspectors were supervised by the relevant public authority.

13.7 Derogation on grounds of public policy, public security or public health in Article 52 TFEU

The ‘public policy’ derogation is rarely invoked, although Calfa (1999), discussed in Chapter 11, involved the application of the public policy derogation to an Italian national convicted of possessing drugs while on holiday in Greece, that is, she was exercising her rights to receive services under Article 56. Omega (Case C–36/02) [2004] ECR I–9609 involved the question of whether or not the protection of human dignity, as provided for in the German Constitution, could be invoked to derogate from the freedom to provide services. The ECJ held that it could.

The Court stated:

An Article 52 derogation was rejected in Navileme & Nautizende (Case C–509/12), unreported. People wishing to sail a boat in Portugal had to have a boating licence, issued following an examination. However, only people resident in Portugal were eligible to sit the exam. Two Portuguese nautical training schools challenged the residency requirement. The Portuguese government claimed that the requirement was necessary on grounds of public policy, in order to ensure a high level of safety at sea. The ECJ held that the requirement infringed the freedom of nautical colleges in Portugal to provide services (in the sense of nautical training services), and the freedom of EU citizens to receive those services, and was prima facie prohibited by Article 56 TFEU. Moreover, the requirement was not justifiable on public policy grounds. First, Article 52 TFEU presupposed the existence of a genuine, sufficiently serious threat affecting one of the fundamental interests of society, which was not the case here. Second, the residence requirement bore ‘no relation’ to the training followed or the ability to sail and hence was ‘not in itself appropriate for attaining the objective in question, that is, to ensure safety of navigation at sea’. Third, the residence condition failed the proportionality test. The objective of ensuring a better level of maritime safety may be satisfied by less restrictive means such as setting the requirements of the examination for the award of a boating licence at a ‘high level’.

The public health derogation is more frequently invoked. For example, in MacQuen & Others (Case C–108/96) [2001] ECR I–837, the ECJ held that a rule of Belgian law which meant that only those holding specific medical qualifications, such as ophthalmologists, had the right to carry out eyesight examinations, was capable of restricting the freedom to provide services. However, the Court went on to hold that the Belgian law was justified on the ground of health protection. Another example is Corporación Dermoestética (Case C–500/06) [2008] ECR I–5785, involving Italian legislation prohibiting the advertising of medical and/or surgical treatments at private health care clinics on national TV. The Court accepted, at least in principle, that such rules ‘can be justified in the light of the objective of protection of public health’.

However, by far the most common situation in which the human health derogation has been invoked is in cases involving individuals receiving private medical treatment in a Member State other than their home state, paying for it themselves, but then seeking to recover the cost of the treatment from their social security provider. Often, the authorities in the home state are very reluctant to meet these costs and seek to justify their refusal on public health grounds. Essentially, the argument is that for one state to have to use public funds to pay for private medical treatment provided in another state would potentially be such an enormous drain on the resources of the former state that it could undermine the ability of that state to provide effective health care treatment itself.

For example, in Kohll (Case C–158/96) [1998] ECR I–1931, the Court stated:

the objective of maintaining a balanced medical and hospital service open to all, although intrinsically linked to the method of financing the social security system, may fall within the derogations on grounds of public health, insofar as it contributes to the attainment of a high level of health protection.

The Court added:

The Court is also very conscious of the enormous amount of centralised planning that has to go into the organisation of hospitals and associated facilities in each Member State. This could be severely disrupted if large numbers of people decided to go abroad to receive private medical treatment instead. Thus, in Watts v Bedford Primary Care Trust (2006), discussed above, involving a British woman who received a hip replacement in a French private hospital, the Court ruled:

There is now secondary legislation in this area. Article 1(1) of Directive 2011/24 (the Cross-border Healthcare Directive) provides: ‘This Directive provides rules for facilitating the access to safe and high-quality cross-border healthcare and promotes cooperation on healthcare between Member States, in full respect of national competencies in organising and delivering healthcare.’ In the context of this chapter, the key provision is Article 7(7), which states that the home Member State may impose on an individual seeking to recover the cost of cross-border healthcare ‘the same conditions, criteria of eligibility and regulatory and administrative formalities … as it would impose if this healthcare were provided in its territory’. However, no such conditions, criteria or formalities:

may be discriminatory or constitute an obstacle to the free movement of patients … unless it is objectively justified by planning requirements relating to the object of ensuring sufficient and permanent access to a balanced range of high-quality treatment in the Member State concerned or to the wish to control costs and avoid, as far as possible, any waste of financial, technical and human resources.

13.8 Justification for non-discriminatory rules

In addition to the specific Treaty derogations, the ECJ has created a parallel set of derogations which may be pleaded by Member States to justify restrictions on establishment or the provision of services.

In Van Binsbergen (1974), the ECJ was asked whether a Dutch rule requiring lawyers to be habitually resident in the Netherlands before they could exercise rights of audience before Dutch courts and tribunals was compatible with Article 56. The ECJ stated:

In Säger v Dennemeyer & Co (1991), the ECJ stated:

The ECJ confirmed the applicability of these criteria in Gebhard (1995) (an Article 49 case). Thus, for a national rule which restricts either the freedom of establishment or the freedom to provide services to be compatible with Article 49 or Article 56, a four-part test has to be satisfied:

  1. the rule must be non-discriminatory;
  2. it must be justified by imperative requirements in the general interest;
  3. it must be suitable for the attainment of the objective it pursues;
  4. it must not go beyond what is necessary in order to attain its objective (the ‘proportionality’ doctrine).

Commission v France (Tourist Guides) (Case C–154/89) [1999] ECR I–659 provides a good example of these conditions being applied. French legislation required all tourist guides accompanying groups of tourists to take an exam in order to become licensed. The Commission alleged that this was in breach of Article 56. France argued that it was necessary to ensure that guides gave tourists correct artistic and cultural information. The ECJ held that the French legislation was a prima facie infringement of Article 56, because it prevented self-employed tour guides from offering their services to tourists; it also prevented tourists from taking part in such organised tours from availing themselves at will of the services in question. However, the Court held that the licence requirement was justifiable, in principle.

  1. It was non-discriminatory, as it applied to French nationals hoping to work as tourist guides as well as to other nationals.
  2. It was at least justifiable, as it sought to ‘ensure the protection of general interests relating to the proper appreciation of places and things of interest and the widest possible dissemination of knowledge of the artistic and cultural heritage of the country’.
  3. It was capable of achieving this objective.
  4. The final question was whether the licence requirement was actually necessary. Here, the Court held that it went too far, at least as far as general tourist information was concerned. Forcing tourist companies to employ only licensed guides would inevitably lead some tour operators to have recourse to local guides instead of their own staff. That, in turn, could leave tourists with a guide unfamiliar with their language, their interests and their specific expectations. Moreover, the fierce competition under which tour companies operated meant that they were obliged voluntarily to exercise control over the tour guides they employed.

In Alpine Investments (1995), the ECJ held that the restriction under Dutch law of ‘cold-calling’ practices in the financial services sector was a prima facie infringement of Article 56. However, it was justifiable in the interests of consumer protection and the protection of the reputation of that service industry. The ECJ concluded with the wry observation that since ‘the commodities futures market is highly speculative and barely comprehensible for non-expert investors, it was necessary to protect them from the most aggressive selling techniques’. The ECJ was also prepared to hold that the Dutch rules were not disproportionate:

  • the legislation prohibited cold-calling only, i.e. it did not prohibit other forms of approaching potential clients;
  • there was no prohibition on contacting existing clients by telephone;
  • the legislation only prohibited cold-calling in the commodities futures market; i.e. brokering in other markets was not subject to the same rules.

Alpine had argued that simply requiring unsolicited telephone calls to be tape-recorded would suffice to protect customers as effectively. Moreover, in the UK, the Securities and Futures Authority had adopted similar rules. However, the ECJ was not persuaded, commenting that the fact that one Member State imposes less strict rules than another does not mean that the latter state’s rules are disproportionate and hence incompatible with EU law.

In Viking Line (Case C–438/05) [2007] ECR I–10779, the ECJ recognised that the protection of workers was a legitimate interest capable of overriding the freedom of establishment.

More recently, the Court acknowledged that environmental protection (long recognised as capable of justifying restrictions on the free movement of goods – see Chapter 14) was also capable of justifying restrictions on the freedom to provide services.

Environmental protection was one of three justifications invoked in Yellow Cab (Case C–338/09) [2010] ECR I–13927. The Court held that Austrian legislation designed to encourage the operation of public transport services was capable of achieving a number of overriding reasons in the public interest: promotion of tourism, road safety (by channelling tourist traffic) and environmental protection (by offering a viable alternative to individual transportation).

In Josemans (Case C–137/09) [2010] ECR I–13019, the Court accepted that a policy introduced in the Dutch city of Maastricht under which non-residents of the Netherlands were barred from entering ‘coffee-shops’ (where cannabis can be bought and consumed alongside food and non-alcoholic drinks in strictly limited quantities and in controlled conditions) in the city was a restriction on the freedom to provide and receive services but was justified on the basis of the need to tackle ‘drug-tourism’ and the accompanying public nuisance. According to the local authorities, before the ban was introduced, the 14 coffee-shops in Maastricht attracted around 10,000 visitors per day (i.e. nearly four million visitors per year), 70% of whom were not resident in the Netherlands. These large numbers of people generated nuisance and crime, in particular dealing in hard drugs.

In FA Premier League & Others; Murphy (Cases C–403, 429/08) [2012] 1 CMLR 29, the Court held that the protection of intellectual property rights, and the protection of sporting events, were overriding reasons in the public interest capable of justifying restrictions on the freedom to provide and receive services (in this case, satellite TV broadcasts of live football matches).

In P.I. (Case C–230/18) (2019), the Austrian police invoked both prevention of crime and health protection as potential justifications for their decision to shut down P.I.’s massage parlour, which they suspected was being used for prostitution. Although prostitution was actually lawful in that region of Austria, advance authorisation was required, and P.I. had not obtained the necessary licence. The police argued that unauthorised prostitution posed a risk of the transmission of disease, human trafficking, forced prostitution and child prostitution. The ECJ accepted these arguments, although the Court went on to say that the police’s powers to shut down the parlour at very short notice and without giving written reasons infringed P.I.’s rights under the Charter (the Article 47 right to an effective remedy and a fair trial).

‘Overriding reasons relating to the public interest’: summary

The case law reveals that a number of different ‘overriding interests’ have been accepted by the ECJ. They include (bear in mind that this is not a comprehensive list – indeed, new interests may be added to the list at any time):

  • consumer protection – Alpine Investments (1995);
  • the ‘sound administration of justice’ – Wouters and Others (Case C–309/99) [2002] ECR I–1577;
  • the need to ‘ensure high standards of university education’ – Neri v European School of Economics (Case C–153/02) [2003] ECR I–13555;
  • the ‘protection of workers’ – Viking Line (2007);
  • the ‘promotion of research and development’ – Jundt (Case C–281/06) [2007] ECR I–12231;
  • environmental protection – Regione Sardegna (2009), Yellow Cab (2010);
  • the need to tackle ‘drug tourism’ – Josemans (2010);
  • the protection of intellectual property (such as copyright) – Murphy (2012);
  • the protection of sporting events – Murphy (2012).

A non-exhaustive list of ‘overriding reasons relating to the public interest’ is also provided in Article 4(8) of the Services Directive 2006/123, all of which are based on cases decided by the ECJ. In that sense the Directive does not add anything new to this area of law.

Regulation of gambling

Several cases have addressed the question of whether Member States can regulate gambling services. In Customs and Excise Commissioners v Schindler (Case C–275/92) [1994] ECR I–1039, the facts of which were given in Chapter 10, the ECJ observed that UK anti-lottery legislation (prior to its abolition by the National Lotteries Act 1993), was capable of infringing Article 56. However, it was justifiable because it pursued the following objectives:

  • to prevent crime and to ensure that gamblers would be treated honestly;
  • to avoid stimulating demand in the gambling sector which has damaging social consequences when taken to excess; and
  • to ensure that lotteries could not be operated for personal or commercial profit but solely for charitable, sporting or cultural purposes.

The ECJ concluded that the UK legislation was more concerned with ‘the protection of the recipients of services and, more generally, of consumers’ and was, therefore, justified. The ECJ was also prepared to accept the ‘maintenance of order in society’ as an alternative public interest reason. The ECJ observed that it was ‘not possible to disregard the moral, religious or cultural aspects of lotteries, like other types of gambling’. After observing that ‘lotteries involve a high risk of crime or fraud … particularly when they are operated on a large scale’, it then declared, slightly pompously, that ‘they are an incitement to spend which may have damaging individual and social consequences’. Consequently, the legislation did not infringe Article 56 ‘in view of the concerns of social policy and of the prevention of fraud which justify it’. Somewhat paradoxically, the ECJ concluded by extolling the virtues of national lotteries, pointing out how they ‘may make a significant contribution to the financing of benevolent or public interest activities such as social works, charitable works, sport or culture’.

Similar outcomes were reached in Läärä (Case C–124/97) [1999] ECR I–6067, concerning Finnish legislation restricting the availability of gaming licences, and in Zenatti (Case C–67/98) [1999] ECR I–7289, concerning Italian law regulating gambling. In the latter case, the ECJ held that the Italian law (prima facie contrary to Article 56 but justifiable) sought to:

  • prevent gambling from being a source of private profit;
  • avoid risks of crime and fraud and the damaging individual and social consequences of the incitement to spend which it represents; and
  • allow gambling only to the extent to which it may be socially useful as being conducive to the proper conduct of competitive sports.

In ANOMAR and Others (Case C–6/01) [2003] ECR I–8621, Portuguese legislation restricted the running of games of chance to casinos within gaming zones created by decree. The Associação National de Operadores de Máquinas Recreativas (ANOMAR), an association of gaming machine operators in Portugal, challenged this as restricting the freedom to provide services (i.e. gambling services to foreign tourists and business people). The ECJ held that the Portuguese legislation was capable of restricting the freedom to provide services, as it prevented operators from providing gambling opportunities outside the gaming zones. However, it was justified. The ECJ referred to the familiar objectives of consumer protection and maintaining order in society.

However, the ECJ refused to apply the usual justifications in Gambelli and Others (Case C–243/01) [2003] ECR I–13031, which involved Italian rules restricting the provision of Internet gambling services to state-run or state-licensed organisations. The Court stated that restrictions on gambling services had to be deployed in ‘a consistent and systematic manner’. This might not be the case where a Member State was providing its own gambling services, either directly or indirectly via a system of exclusive licences, whilst prohibiting private operators from doing likewise. The ECJ stated:

The ECJ did not maintain this strict position for long. A partial retreat from Gambelli was announced in Placanica (Cases C–338, 359 and 360/04) [2007] ECR I–1891, when the Court stated that a distinction had to be drawn between the objectives of reducing gambling opportunities on one hand and crime prevention on the other. By focusing on the latter objective, the Court held that a ‘policy of controlled expansion in the betting and gaming sector’ carried out under state control or licence may be ‘entirely consistent with the objective of drawing players away from clandestine betting and gaming … to activities which are authorised and regulated’. The Court declared that a licensing system may ‘constitute an efficient mechanism enabling operators active in the betting and gaming sector to be controlled with a view to preventing the exploitation of those activities for criminal or fraudulent purposes’.

The Placanica principle has been applied in numerous cases subsequently to uphold the situation in several Member States whereby certain forms of gambling are allowed but only under state control (whether the gambling services were provided by the state itself or by private operators under state licence). Although this restricts the freedom of gambling providers based in other states to provide their services, it is justifiable on the basis that state-controlled gambling services are safer for the consumer than equivalent services provided on the open market (see e.g. Betfair (Case C–203/08) [2010] ECR I–4695 (the Netherlands); Sjöberg & Gerdin (Cases C–447 and 448/08) [2010] ECR I–6921 (Sweden); and Zeturf (Case C–212/08) [2011] ECR I–5633 (France)).

A similar outcome to that in ANOMAR occurred in Liga Portuguese de Futebol Profissional (Case C–42/07) [2009] ECR I–7633, also known as the Santa Casa case. Portuguese legislation conferred an exclusive right on an organisation called Santa Casa to operate lotteries and all forms of sports betting (including online gambling) throughout Portugal. The Court held that, although clearly capable of restricting both the freedom of rival companies based in other Member States to provide gambling services and the freedom of Portuguese consumers to receive those services, the Portuguese legislation was justified. In giving judgment, the ECJ focused on crime prevention, stating:

Berlington & Others v Hungary (Case C–98/14) [2015] 3 CMLR 45 involved Hungarian legislation banning the operation of slot machines except in casinos. The claimants operated slot machines in amusement arcades in Hungary. They challenged this legislation, alleging a breach of Article 56, on the basis that most of the people who used arcade machines were tourists from other Member States. The ECJ agreed that the legislation was a restriction on the freedom to provide services but was potentially justifiable on the basis that ‘the protection of consumers against gambling addiction and the prevention of crime and fraud linked to gambling, constitute overriding reasons in the public interest capable of justifying restrictions on gambling’, subject to the proportionality test. The Hungarian legislation would be appropriate for ensuring attainment of the objective relied on only if it reflected a concern to attain it in a ‘consistent and systematic’ manner, a matter for the national court to determine.

Internet gambling

In Santa Casa, the ECJ also held that, because of the particular dangers associated with internet gambling, Member States were not obliged to take into account the fact that the provider of a gambling service might be regulated by the national authorities in another Member State. The justification proffered for this was ‘the lack of direct contact between consumer and operator’, which meant that ‘games of chance accessible via the internet involve different and more substantial risks of fraud by operators against consumers compared with the traditional markets for such games’. In other words, Internet gambling was singled out for special treatment because of the greater threat of criminal activity.

In subsequent cases, the Court has confirmed that Internet gambling is special, but for other reasons than the risk of crime and fraud. In Carmen Media (Case C–46/08) [2010] ECR I–8149, the Court focused on the need to protect against gambling addiction. The Court held that the specific ‘characteristics’ of Internet gambling may prove to be a ‘source of risks of a different kind and a greater order’ compared with traditional forms of gambling, particularly in relation to ‘young persons’ and those with ‘a propensity for gambling or likely to develop such a propensity’. Apart from the lack of direct contact, the Court identified a number of other specific risks, namely the ‘particular ease and the permanence of access to games offered over the internet’ and the ‘potentially high volume and frequency’ of international offers (a point A-G Mengozzi had made in his Opinion, i.e. the supply of games available online is virtually unlimited – a number of gambling windows can be opened at the same time – and it is available 24 hours a day, 365 days a year). These risks, moreover, were exacerbated by the fact that Internet gambling took place ‘in an environment … characterised by isolation of the player, anonymity and an absence of social control’.

In Zeturf (2011), the Court confirmed both Santa Casa and Carmen Media, namely that Internet gambling posed ‘different and more substantial risks of fraud’ and ‘may prove to be a source of risks of a different kind and a greater order in the area of consumer protection’.

Soon afterwards, in Dickinger & Ömer (Case C–347/09) [2011] ECR I–8185, the ECJ again confirmed its policy of applying special rules to Internet gambling. The result was that the authorities in Austria did not have to recognise the controls imposed on the activities of Internet gambling providers established in Malta. The Maltese government, intervening in the case, had sought to distinguish Santa Casa (2009) on the basis that ‘Malta was the first State to develop a regulatory system specifically aimed at controlling and monitoring online games of chance [and] which was designed with the objective of addressing the risks inherent in those modern modes of operation’. Specifically, Internet gambling providers based in Malta were subject ‘to continued checks and monitoring’ by the Lotteries and Gaming Authority, which had ‘implemented advanced and robust systems of regulation’. It was further argued that, if anything, Internet gambling services can be controlled ‘more effectively’ than traditional gambling services ‘because all operations performed on electronic media can be tracked, which makes it easy to detect problematic or suspicious operations’. However, none of these arguments persuaded the Court to change its policy. The Court explained that in the absence of EU-level harmonisation and in view of the ‘substantial differences between the objectives pursued and the levels of protection sought by the legislation of the various Member States’, it was not possible to compel the Member States to recognise each other’s regulatory systems.

13.9 The free movement of lawyers

Lawyers are called a variety of names in the EU, as the following list illustrates:

  • abogado (used in Spain);
  • advocaat (used in the Netherlands and parts of Belgium);
  • advocat or Advokat (used in Denmark, Finland, Slovakia and Sweden);
  • advogado (used in Portugal);
  • advokáat (used in the Czech Republic);
  • advokāts (used in Lithuania);
  • adwocat (used in Poland);
  • avocat (used in France, Luxembourg and parts of Belgium);
  • avukat (used in Malta);
  • avvocato (used in Italy);
  • barrister/solicitor (used in Ireland and the UK);
  • dikigoros (used in Cyprus and Greece);
  • odvetnik/odvetnica (used in Slovenia);
  • rechtsanwalt (used in Austria, Germany and parts of Belgium);
  • ügyvéd (used in Hungary);
  • vandeadvokaat (used in Estonia);
  • zvērināts (used in Latvia).

13.9.1 Freedom of establishment under Article 49 TFEU

Many lawyers, particularly barristers, are self-employed. This means that there have been several cases involving the free movement of lawyers under Article 49. In Paris Bar Council v Klopp (1984), the ECJ was asked to rule on the compatibility of a rule of French law that said that lawyers could have only one place of establishment at a time. This was allegedly designed to ensure that lawyers ‘should practise in such a way as to maintain sufficient contact with their clients and the judicial authorities and abide by the rules of the profession’. The ECJ held that this rule imposed an unjustifiable restriction on the freedom of lawyers to establish themselves in other EU Member States.

The Court stated:

However, national legal authorities may be able to justify other restrictions, such as supervision and compliance with professional rules. Gullung v Colmar and Saverne Bar Council (Case 292/86) [1988] ECR 111 involved a German lawyer who had been refused permission to practise in France after contravening French regulations relating to the professional ethics of those in the legal professions. The Court stated:

In what is now the leading case, Gebhard v Milan Bar Council (1995), the ECJ confirmed Gullung. The ECJ acknowledged that national professional regulatory bodies (such as the Solicitors Regulation Authority and the Bar Standards Board in England and Wales) have rules governing issues such as client care, confidentiality and professional ethics which are capable of imposing restrictions on the free movement of lawyers. Nevertheless, these rules are designed to protect clients and the reputation of the profession as a whole and, therefore, are justifiable.

The Court stated:

13.9.2 The Lawyers’ Establishment Directive 98/5

Directive 98/5 confers rights on lawyers qualified in one Member State (the home state) to practise in another Member State (the host state), although they must be clearly ‘badged’ as a sign to potential clients that the migrant lawyer is qualified in another state’s law.

Practice under the ‘home country professional title’

According to Article 2, migrant lawyers are entitled to practise various activities which are listed in Article 5, but only using their original title, in the host state’s language. According to Article 4, this title ‘must be expressed in the official language … of his home Member State, in an intelligible manner and in such a way as to avoid confusion with the professional title of the host Member State’. Under Article 3, migrant lawyers are required to ‘register with the competent authority’ in the host Member State. The areas of activity in Article 5(1) include advice on:

  • the law of the home Member State;
  • EU law;
  • international law; and
  • the law of the host Member State.

In Torresi (Cases C–58 and 59/13) [2015] QB 331; [2015] 2 WLR 29, the Court was asked whether Directive 98/5, in particular Article 3, applied in the case where a national of one Member State (Italy) left that state in order to acquire the qualifications necessary to practise as a lawyer in another Member State (Spain) before returning to the home state in order to practise there. The Court answered ‘yes’. The Court held that the right of EU citizens to choose, on the one hand, the Member State in which they wished to acquire their professional qualifications and, on the other, the Member State in which they intended to practise their profession was ‘inherent in the exercise, in a single market, of the fundamental freedoms guaranteed by the Treaties’. Moreover, the fact that a lawyer had chosen to acquire a professional qualification in a Member State other than their home state in order to benefit from ‘more favourable’ legislation in that other state was not, in itself, sufficient ground to conclude that there is an abuse of EU rights (which could have precluded the application of Directive 98/5).

Rules of professional conduct

When practising in the host state, even under his home state’s professional title, the migrant lawyer will be expected to abide by ‘the local codes of conduct in respect of all activities he pursues in its territory’ (Article 6). This will extend to, for example, rules on advertising.

Disciplinary proceedings

The migrant lawyer will be subject to all the local disciplinary rules (Article 7(1)). Provision is made for his home state’s professional body to be fully informed of any disciplinary action, including ‘all the relevant details’, i.e. the evidence on which it is based, before any action is taken (Article 7(2)). A duty of co-operation is imposed on the host state’s disciplinary body, while the home state’s body has a right to ‘make submissions to the bodies responsible for hearing any appeal’ (Article 7(3)).

The registration requirement in Article 3, as well as the professional conduct and disciplinary proceedings requirements in Articles 6 and 7, of Directive 98/5 were all examined in Wilson (Case C–506/04) [2006] ECR I–8613, involving a British barrister, Graham Wilson, who wished to register in Luxembourg. The Court held that Article 3 did not allow the registration of a lawyer in a Member State to be made conditional on them being proficient in the language(s) of that state. The Court stated: ‘presentation … of a certificate attesting to registration with the competent authority of the home State is the only condition to which registration of the person concerned in the host Member State may be subject’. However, the Court noted that, under Article 6, a lawyer must comply with the rules of professional conduct applicable in the host state. One of those rules is an obligation not to handle matters which a lawyer knows, or ought to know, they are not competent to handle, for example, because they lack linguistic knowledge. The Court held that communication with clients, with the national authorities and/or with the professional bodies of the host state may all require a lawyer to have sufficient linguistic knowledge, or at least recourse to assistance where that knowledge is insufficient. Acting without such knowledge may amount to a breach of Article 6 of the Directive leading to possible disciplinary action under Article 7. Nevertheless, the Court concluded, the linguistic knowledge requirement did not justify the national authorities demanding an examination of the lawyer’s proficiency in the language(s) of the host Member State prior to registration.

Admission to the legal profession in the host state

Article 10(1) provides that a lawyer practising under his home country professional title who has ‘effectively and regularly’ pursued an activity in the host Member State in the law of that state (including EU law) for at least three years shall, with a view to gaining admission to the profession of lawyer in the host Member State, be exempted from the conditions set out in (what is now) Article 14(1) of Directive 2005/36 (the Qualifications Directive).

Article 10 was invoked in Ebert (Case C–359/09) [2011] ECR I–269. Donat Ebert, a German lawyer, had practised law in Hungary under his home country title, ‘rechtsanwalt’, for several years. He applied to the Hungarian authorities for permission to use the Hungarian title, ‘ügyvéd’. This was rejected because Ebert had not become a member of the Hungarian Bar Association. He challenged this refusal. However, the Court held that Article 10 exempted certain lawyers from the need to comply with Directive 2005/36; it did not preclude national rules laying down a requirement to be a member of a body such as a national Bar Association in order to practise law under the title of the host Member State.

13.9.3 The Lawyers’ Services Directive 77/249

Directive 77/249 harmonises the rules under which lawyers established in one Member State may provide their services in the territory of another state. Article 2 provides: ‘Each Member State shall recognise as a lawyer for the purpose of pursuing the activities specified in Article 1(1) any person listed in Article 1(2)’. This is the same list as found in Directive 98/5. Article 1(1) simply states that ‘This directive shall apply, within the limits and under the conditions laid down herein, to the activities of lawyers pursued by way of provisions of services.’

In Piringer (Case C–342/15) [2017] 3 CMLR 19, the ECJ held that Directive 77/249 covers the situation where the lawyer travels to another Member State to meet with their client and provide legal services there and to the corollary situation where the client travels to another Member State to meet with their lawyer and receive legal services there. Ms Leopoldine Piringer, an Austrian national, had travelled to the Czech Republic to meet with her lawyer in order to have her signature on a document in which she had declared her intention to sell property that she owned in Austria authenticated. The Court held that Directive 77/249 was ‘capable of applying’ on these facts. (Ultimately, Ms Piringer was barred from using her Czech lawyer’s services in this particular case but only because of a specific derogation in the directive which provides that Member States have the option of reserving to ‘prescribed categories of lawyers’ the preparation of formal documents for, inter alia, creating or transferring rights to property.)

SUMMARY

  • Article 49 TFEU covers the freedom of establishment; Article 56 TFEU covers the freedom to provide services. Both are directly effective: Reyners (Article 49) and Van Binsbergen (Article 56).
  • Establishment is the right to install oneself in another Member State, permanently or semi-permanently, on a self-employed basis, for the purpose of performing a particular activity there. It also gives companies the right to set up a branch or a subsidiary, in another Member State.
  • Provision of services allows an individual, established in one Member State, to provide their services in another Member State, on a temporary or intermittent basis. It also allows a company, established in one Member State, to provide their services to anyone in another Member State. If necessary, it allows them to visit the other Member State, on a temporary or intermittent basis, in order to do so.
  • Article 49 refers to the taking-up as well as the pursuit of professional activities for individual, self-employed persons. For companies, the freedom of establishment includes the right to set up and manage ‘undertakings’. In both cases, establishment is under the same conditions laid down for nationals of the state where establishment is effected (Steinhauser).
  • It is possible to be established in two Member States at the same time (Klopp, Gebhard).
  • Article 49 can be invoked vertically (Reyners) and horizontally (Viking Line).
  • Directive 2005/36 (as amended) consolidates the rules for the mutual recognition of qualifications required in order to pursue a ‘regulated profession’ in another Member State.
  • A profession is ‘regulated’ if access to it depends on the possession of qualifications (Aranitis).
  • Once qualifications have been recognised, the holder of them can pursue a profession in the host Member State under the same conditions as its nationals.
  • Member States are allowed to impose ‘compensation measures’ on holders of qualifications obtained in other states in certain situations, e.g. if the duration of training is at least one year shorter than in the host state, or if training covers ‘substantially different matters’.
  • Where that happens, states may require the completion of an ‘adaptation period’ or the passing of an ‘aptitude test’. Usually, the candidate can choose which, except for professions ‘whose pursuit requires precise knowledge of national law’, in which case the state can specify. This exception does not just apply to ‘traditional’ legal professions (Price, Vandorou & Others). Alternatively, states can offer ‘partial recognition’ of a candidate’s qualifications (Colegio de Ingenieros, Nasiopoulos).
  • Where Directive 2005/36 does not apply, states are obliged to undertake an assessment of a candidate’s qualifications to establish the extent of equivalence, which may be full, partial or not equivalent (Thieffry, Patrick, Vlassopoulou, Morgenbesser, B.B., A).
  • Where qualifications obtained outside the EU have been accepted by one Member State as equivalent to their qualifications and the candidate has practised there, other states must recognise that practical experience (Haim, Hocsman).
  • Article 56 TFEU provides that ‘restrictions on freedom to provide services … shall be prohibited’. It is crucial that there is a cross-border element. Article 56 EC applies if two conditions are satisfied: (1) the service must be provided within the EU; (2) the provider of services must be a national of a Member State and established in a state of the EU other than that of the service recipient (Scorpio Konzertproduktionen).
  • Article 56 may be invoked by service providers to challenge all national legislation, whether discriminatory or not, if it inhibits the freedom to provide services (Säger v Dennemeyer & Co).
  • Article 56 may be used by service providers to challenge legislation in their state of establishment (Alpine Investments, Carpenter).
  • Article 56 may be invoked by the recipients of services, e.g. tourists (Luisi & Carbone, Calfa), persons receiving private education (Humbel, Schwarz), persons receiving medical treatment (Kohll, Watts).
  • ‘Services’ has a very wide scope, including TV broadcasting (Bond van Adverteerders), gambling (Schindler), abortions (SPUC v Grogan), prostitution (Jany and Others), sporting activities (Deliège).
  • The service must normally be provided for ‘remuneration’ (Article 57 TFEU). Thus, while private education is included, public education is not (Humbel, Schwarz). It is not essential that the service recipient be the person who provides the remuneration (Bond van Adverteerders, Deliège).
  • Establishment and services are mutually exclusive; the freedom to provide services is subordinate to the freedom of establishment; the concept of establishment assumes the person/company in question is in another Member State ‘on a stable and continuous basis’, whereas the provisions on services ‘envisage that he is to pursue his activity there on a temporary basis’ (Gebhard).
  • Ultimately, whether Article 49 or 56 applies is ‘determined in the light, not only of the duration of the provision of the service, but also of its regularity, periodicity or continuity’ (Gebhard).
  • ‘Activities … that are connected, even occasionally, with the exercise of official authority’ are not covered by the Treaty (Article 51 TFEU). This applies only to official authority activities; not professions or occupations as a whole (Reyners).
  • Member States may derogate from the freedoms on grounds of public policy, security or health (Article 52 TFEU). For further explanation, see Chapter 11.
  • Article 52 allows Member States, in principle, to restrict access to medical services on public health grounds. This is justifiable because it ‘seeks to ensure that there is sufficient and permanent access to a balanced range of high-quality hospital treatment in the State concerned’ (Kohll, Watts).
  • National legislation restricting either of the freedoms is justifiable if it is non-discriminatory, pursues a legitimate objective, is capable of achieving that objective and satisfies proportionality, i.e. it does not go beyond what is necessary to achieve that objective (Van Binsbergen, Gebhard).
  • Examples of legitimate objectives include ‘the proper appreciation of places and things of historical interest; dissemination of knowledge of national artistic and cultural heritage’ (Commission v France (Tourist Guides)), consumer protection (Alpine Investments, Schindler), the protection of workers (Viking Line), protection of the environment (Regione Sardegna, Yellow Cab), the prevention of crime (Zenatti, Santa Casa).
  • In addition, Article 16 of Directive 2006/123 (the Services Directive), states that Member States must not make access to or exercise of a service activity in their territory subject to compliance with any requirements which do not respect the principles of non-discrimination, necessity and proportionality.
  • Lawyers seeking establishment in another Member State may rely on Directive 98/5, while lawyers providing services may rely on Directive 77/249, in addition to the rights provided by Articles 49 and 56.

Further reading

Articles

  • Anagnostaras, G, ‘Les Jeux Sont Faits? Mutual Recognition and the Specificities of Online Gambling’ (2012) 37 EL Rev 191.
  • Barnard, C, ‘Unravelling the Services Directive’ (2008) 45 CML Rev 323.
  • Bulterman, M and Kranenborg, H, ‘What if Rules on Free Movement and Human Rights Collide? About Laser Games and Human Dignity: The Omega Case’ (2006) 31 EL Rev 93.
  • Davies, G, ‘The Services Directive: Extending the Country of Origin Principle, and Reforming Public Administration’ (2007) 32 EL Rev 232.
  • De La Rosa, S, ‘The Directive on Cross-Border Healthcare or the Art of Codifying Complex Case Law’ (2012) 49 CML Rev 15.
  • Hatzopoulos, V, ‘The Court’s Approach to Services (2006–2012): From Case Law to Case Load?’ (2013) 50 CML Rev 459.
  • Littler, A, ‘Regulatory Perspectives on the Future of Interactive Gambling in the Internal Market’ (2008) 33 EL Rev 211.
  • Van den Bogaert, S and Cuyvers, A, ‘“Money for Nothing”: The Case Law of the EU Court of Justice on the Regulation of Gambling’ (2011) 48 CML Rev 1175.

14 The free movement of goods and Articles 34 and 35 TFEU

DOI: 10.4324/9781003218562-14

14.1 The removal of non-fi scal barriers to trade

In Chapter 15 we will examine how EU law has sought to prevent Member States from imposing fiscal barriers (in the forms of Customs charges and discriminatory taxation) to the free movement of goods. In this chapter, we will examine how EU law prevents Member States from imposing non-fiscal barriers. There is a potentially infinite variety of such barriers, from measures such as hygiene inspections at border crossing points to technical legislation prescribing the permitted amount of salt in bakery products or the shape of margarine tubs. National legislation making it a criminal offence to import pornographic videos or banning the advertising of junk food could all have the effect of imposing trade barriers.

14.2 Prohibition of quantitative restrictions on imports (Article 34) and exports (Article 35) and all measures having equivalent effect

14.2.1 Introduction

Article 34 prohibits quantitative restrictions, and all measures having equivalent effect, on imports; Article 35 does the same for exports. A measure which infringes Articles 34 or 35 is prima facie contrary to EU law; however, Article 36 provides that Articles 34 or 35 will not apply to certain restrictions, justifiable on various grounds, which are not disproportionate (Article 36 is discussed in section 14.5; proportionality is discussed in section 14.7). In addition to this, the ECJ has developed its own line of case law, allowing justifications on other grounds, again provided that they are not disproportionate (see Cassis de Dijon (Case 120/78) [1979] ECR 649, discussed in section 14.6).

14.2.2 Direct effect of Articles 34 and 35

In Ianelli & Volpi SpA v Meroni (Case 74/76) [1977] ECR 595, the ECJ announced:

14.2.3 Scope of Articles 34 and 35

Articles 34 and 35 are addressed to the Member States and therefore apply only to acts or omissions on behalf of the Member States. This essentially means the legislative and executive arms of each state’s government although it extends beyond that. In Aragonesa and Publivia (Cases C–1 and 176/90) [1991] ECR I–4151, which concerned provisions of Catalan law, the ECJ stated that Article 34 ‘may apply to measures adopted by all the authorities of the Member States, be they the central authorities, the authorities of a federal State, or other territorial authorities’. Apart from central and local government, the actions of the following have been held capable of infringing Articles 34 and 35:

  • semi-public bodies such as quangos (e.g. Apple and Pear Development Council v K Lewis Ltd) (Case 222/82) [1983] ECR 4083;
  • nationalised industries such as the Post Office (Commission v France (Case 21/84) [1985] ECR 1355);
  • regulatory agencies and professional bodies established under statutory authority (R v Pharmaceutical Society of GB, ex parte Association of Pharmaceutical Importers (Cases 266 and 267/87) [1989] ECR 1295);
  • the police force (R v Chief Constable of Sussex, ex parte ITF Ltd [1998] 3 WLR 1260, HL);
  • even the EU’s own institutions are bound to comply with the provisions of Articles 34 and 35 (Denkavit (Case 15/83) [1984] ECR 2171).

The word ‘goods’ does not appear in either Article 34 or Article 35. The word does appear in Article 36, however, which refers to ‘prohibitions or restrictions on imports, exports or goods in transit’. It also appears in Article 26(2) and Article 28(1) TFEU, and it has been defined – very widely – by the ECJ (refer to Chapter 10, section 10.2.3). In practice, all products (whether manufactured or not) taken across an internal border of the EU for the purpose of commercial transactions are subject to the free movement of goods provisions – even waste (Commission v Belgium (Case C–2/90) [1992] ECR I–4431; Dusseldorp (Case C–203/96) [1998] ECR I–4075). The provisions apply equally to goods manufactured or produced in the EU and to those in ‘free circulation’ in the EU, regardless of their country of origin. Thus, in Donckerwolcke and Schou (Case 41/76) [1976] ECR 1921, where cloth originating from Syria and the Lebanon had been imported into Belgium before being re-imported into France, the ECJ held that Article 34 applied.

14.3 The definition of ‘quantitative restrictions’

A ‘quantitative restriction’ was defined in Geddo v Ente Nazionale Risi (Case 2/73) [1973] ECR 865 as ‘measures which amount to a total or partial restraint of, according to the circumstances, imports, exports or goods in transit’. This most obviously includes a quota system (Salgoil (Case 13/68) [1968] ECR 453), but also includes an outright ban on imports (Commission v Italy (Case 7/61) [1961] ECR 635; Commission v UK (Case 40/82) (the French Turkeys case) [1982] ECR 2793). In R v Henn and Darby (Case 34/79) [1979] ECR 3795, the ECJ held that s42 of the UK’s Customs Consolidation Act 1876, which made it a criminal offence to be ‘knowingly concerned in the fraudulent evasion of the prohibition of the importation of obscene articles’, was a quantitative restriction. The ECJ stated:

This was contrary to what had been earlier suggested in the Court of Appeal in the same case, by Lord Widgery CJ, that a total prohibition was not a ‘quantitative’ restriction because the 1876 Act made no reference to quantities of obscene articles ([1978] 3 All ER 1190).

Rosengren & Others (Case C–170/04) [2007] ECR I–4071 concerned Swedish legislation which prohibited individuals from importing spirits, wine or strong beer into Sweden, unless personally transporting it. A number of Swedish nationals had ordered cases of Spanish wine to be imported into Sweden. However, the wine was confiscated by Swedish customs. A legal challenge was brought to recover the wine, which raised a question regarding the legality of the Swedish legislation under EU law. The ECJ held that the Swedish legislation constituted a quantitative restriction on imports.

Generally speaking, it will be positive actions that infringe Articles 34 and 35. However, it will be possible for Member States to infringe Article 34, at least, by omission. This has occurred in two cases:

  • Commission v France (French Farmers) (Case C–265/95) [1997] ECR I–6959, where the French authorities failed to take action to prevent striking French farmers from blockading ports.
  • Schmidberger v Austria (Case C–112/00) [2003] ECR I–5659, where the Austrian authorities decided to allow a demonstration by an environmental group to go ahead. The effect was to block a major motorway to heavy goods vehicles.

In both cases the French and Austrian authorities, respectively, were held to have infringed Article 34 – although in Schmidberger (2003) the ECJ went on to decide that the authorities’ omission was justified (this case is considered below – see section 14.6).

14.4 Defining ‘measures equivalent to quantitative restrictions’ (MEQRs) in Article 34: the Dassonville formula

Nowhere in the TFEU is the phrase ‘measures equivalent to quantitative restrictions’ (known as MEQRs) defined. The classic formulation was given in Dassonville (Case 8/74) [1974] ECR 837, a case involving a provision of Belgian law found to amount to an MEQR.

The Court stated:

This definition of an MEQR has been cited in practically every case involving Article 34 ever since. However, in some of the more recent cases, the ECJ has modified the ‘formula’ slightly, substituting the word ‘commercial’ for ‘trading’.

Note that the ‘formula’ extends Article 34 to any measure that might affect trade (as well as measures that definitely or probably would affect trade, or have actually done so). Dassonville (1974) therefore gives Article 34 a very wide scope indeed, as the cases examined below illustrate. But there are some limitations. A ‘charge having equivalent effect to a Customs duty’ cannot also be a measure equivalent to a quantitative restriction – that is, a measure cannot be in breach of both Article 30 and Article 34 (Ianelli & Volpi v Meroni (1977)). (Article 30 is examined in Chapter 15.)

14.4.1 Distinctly applicable MEQRs

In Dassonville (1974), the ECJ did not distinguish between those national rules which only apply to imports (known as ‘distinctly applicable’ MEQRs), and those national rules which apply both to imports and domestically produced goods (known as ‘indistinctly applicable’ MEQRs). However, the distinction between the two types of national rules is very important, as it determines whether or not the Cassis de Dijon principle (discussed below) applies.

The following are some examples of ‘distinctly applicable’ MEQRs:

‘Buy national’ campaigns

Government-sponsored campaigns to encourage consumers to buy domestic products on the basis of their nationality clearly infringe Article 34. See Commission v Ireland (‘Buy Irish’ Campaign) (Case 249/81) [1982] ECR 4005:

However, it is not contrary to EU law to promote a domestic product by pointing out that it has certain qualities not found in goods from other Member States (Apple and Pear Development Council v K Lewis Ltd (1983)):

Import licence requirements

Where national legislation insists on importers being licensed, the ECJ has held that such requirements are in breach of Article 34. Evans Medical & Macfarlan Smith (Case C–324/93) [1995] ECR I–563 concerned licences to import poppy seeds into the UK. The British government insisted that importers be licensed, because although poppy seeds can be converted legally into diamorphine, a powerful pain-killing drug widely used in British hospitals, if it falls into the ‘wrong hands’ it can end up on ‘on the street’ as a heroin substitute. Article 34 applies even if the granting of the licence would be a mere formality (Commission v UK (UHT Milk) (Case 124/81) [1983] ECR 203). This is because the cost and time taken up in having to apply for a licence could act as a barrier to trade (Franzén (Case C–189/95) [1997] ECR I–5909).

Similarly, national legislation requiring retailers to be licensed in order to sell a product has also been held to breach Article 34 (Sandoz (Case 174/82) [1983] ECR 2445; Visnapuu (Case C–198/14) [2016] 2 CMLR 32). However, this particular type of licensing system may now be regarded, after the decision in Keck and Mithouard (Case C–267/91) [1993] ECR I–6097, as a selling arrangement, which is exempt from Article 34.

Hygiene inspections

Hygiene inspections carried out on imported goods (typically food and drugs) may still infringe Article 34 because they involve delay, expense, etc. This was seen in the following cases:

  • Rewe-Zentralfinanz (Case 4/75) [1975] ECR 843 (German legislation required that imported apples be subject to phytosanitary inspection (to detect the presence of San José scale));
  • Commission v France (Italian Wines) (Case 42/82) [1983] ECR 1013 (French legislation required imported wine from Italy to be subjected to rigorous inspections).

14.4.2 Indistinctly applicable MEQRs

The majority of the cases have involved indistinctly applicable MEQRs, that is, national rules which apply without distinction to imports and to domestically produced goods but which, nevertheless, have the potential to hinder trade.

Origin marking requirements

National laws imposing a requirement that goods be marked with their country of origin could infringe Article 34 for two reasons:

  • They impose extra burdens on importers, many of whom will also not even be aware of the national law and so face difficulties in complying with it.
  • They may encourage ‘latent’ nationalistic prejudice in shoppers, who may consciously or subconsciously select domestically produced goods in preference to imports, purely on the basis of their nationality.

In Commission v UK (Origin Marking) (Case 207/83) [1985] ECR 1201, the ECJ acknowledges both of these arguments:

Other examples include Dassonville (1974) itself and Commission v Ireland (Souvenir Jewellery) (Case 113/80) [1981] ECR 1625. The Souvenir Jewellery case will be discussed below.

Packaging requirements

National laws which relate to how products are packaged may well infringe Article 34, because they increase the costs of manufacturers in other Member States, who will have to develop special packaging processes purely for the importing state. It will also inhibit retailers in the state in question from importing goods that do not comply with the national law. Conversely, it will be much easier for domestic manufacturers to comply with their own national requirements as to packaging. Examples include:

  • Walter Rau v De Smedt (Case 261/81) [1982] ECR 3961 (Belgian legislation required margarine to be packaged in a cube);
  • Mars (Case C–470/93) [1995] ECR I–1923 and Estée Lauder Cosmetics v Lancaster (Case C–220/98) [2000] ECR I–117 (German legislation prohibited the use of misleading packaging).

Contents and ingredients restrictions

In the following cases, national legislation prescribing or restricting the contents and/or ingredients of various products was held to breach Article 34:

  • Cassis de Dijon (Case 120/78) [1979] ECR 649: German legislation laid down a minimum alcohol level of 25% per litre for certain spirits.
  • Gilli and Andres (Case 788/79) [1981] ECR 2071: Italian legislation required all vinegar to be made from wine.
  • Commission v Germany (Beer Purity) (Case 178/84) [1987] ECR 1227: German legislation prohibited the use of additives in beer.
  • Muller (Case 304/84) [1986] ECR 1511: French legislation prohibited the use of emulsifying agents in bakery products.
  • Greenham and Abel (Case C–95/01) [2004] ECR I–1333: French legislation prohibited the sale of any food or drink containing a chemical substance called coenzyme Q10.
  • Commission v Italy (Red Bull) (Case C–420/01) [2003] ECR I–6445: Italian legislation banned drinks with more than 125 mg per litre of caffeine (such as ‘Red Bull’, which has a caffeine level double that).

Many of these cases reached the ECJ via requests for preliminary rulings from criminal courts because traders had been prosecuted for selling imported goods that did not comply with the national legislation. Thus:

  • Herbert Gilli and Paul Andres were prosecuted in Italy for selling apple vinegar made in Germany.
  • Claude Muller was prosecuted in France for selling a cake and pastry mix called ‘Phénix’, imported from Germany, which contained an emulsifying agent.
  • John Greenham and Léonard Abel were prosecuted in France for selling a food supplement, ‘Juice Plus’, to which had been added Q10. ‘Juice Plus’ is sold without restriction in Germany, Italy and Spain.

In all of these cases the ECJ decided that the national rules in question amounted to an indistinctly applicable MEQR, and were thus prohibited by Article 34, unless one of the Article 36 or Cassis de Dijon (1979) derogations applied (typically, protection of public health).

Name restrictions

National legislation that reserves particular names to products bearing very specific characteristics is capable of breaching Article 34. For example, in Smanor (Case 298/87) [1988] ECR 4489 (French legislation reserved the name ‘yoghurt’ to fresh produce only, with the result that frozen yoghurt had to be re-named as ‘deep-frozen fermented milk’) the ECJ decided that Article 34 had been infringed because ‘it may none the less make the marketing [of imported frozen yoghurt] more diffi-cult and thus impede, at least indirectly, trade between Member States’.

The same result occurred in the following cases:

  • Fietje (Case 27/80) [1980] ECR 3839: Dutch legislation made the name ‘likeur’ compulsory for most alcoholic products of at least 22% proof.
  • Miro (Case 182/84) [1985] ECR 3731: Dutch legislation prescribed that the word ‘Jenever’ could only be applied to describe gin that was at least 35% proof.
  • Deserbais (Case 286/86) [1988] ECR 4907: French legislation restricted the use of the word ‘Edam’ to describe cheese with a minimum fat content.
  • Guimont (Case 448/98) [2000] ECR I–10663: French legislation prescribed the contents of ‘Emmenthal’ cheese very rigidly, in that it had to be a firm cheese produced by curing, pressing and salting on the surface or in brine, of a colour between ivory and pale yellow, with holes of a size between a cherry and a walnut [and with a] hard, dry rind, of a colour between golden yellow and light brown.
  • Commission v Spain (Case C–12/00) [2003] ECR I–459: Spanish and Italian legislation restricted the use of the name ‘chocolate’ to products containing only chocolate and no vegetable fats. This affected chocolate products made in the UK, Denmark, Finland, Ireland, Portugal and Sweden, which traditionally contain vegetable fats. These could be sold in Italy and Spain but only under the label ‘chocolate substitute’.

Conversely, where national authorities in one Member State ban or restrict the use of a name which is used elsewhere, this could also constitute a breach of Article 34. Thus, in Clinique Laboratories and Estée Lauder Cosmetics (Case C–315/92) [1994] ECR I–317, where the German authorities refused to allow the name ‘Clinique’ to be used for cosmetics, the ECJ held that Article 34 had been infringed.

Authorisation/certification requirements

National legislation requiring all goods of a certain type to be inspected in order to ensure that they satisfy national standards, and authorised or certified as such prior to them being made available for sale to the consumer, are capable of hindering trade and are therefore in breach of Article 34. Examples include:

  • Dynamic Medien v Avides Media (Case C–244/06) [2008] ECR I–505 – German legislation required DVDs to be inspected and classified.
  • Commission v Spain (Herbal Products) (Case C–88/07) [2009] ECR I–1353 – under Spanish legislation, all medical products could only be placed on the market with official authorisation.
  • ASCAFOR & ASIDAC (Case C–484/10) [2012] 2 CMLR 22 – under Spanish legislation, a number of technical requirements had to be met before reinforced steel could be used in construction projects.
  • Fra.bo (Case C–171/11) [2012] 3 CMLR 38; [2013] QB 187 – under German legislation, only products (the case itself involved copper fittings for use in the water industry) which met certain technical standards could be used in Germany.

In such cases, although a prima facie breach of Article 34 was identified (on the basis that the national requirements posed a potential barrier to trade), the legislation was justifiable (in principle), typically on health and/or consumer protection grounds.

Prohibitions on use

A complete ban under national legislation on the use of a product is an MEQR. In Toolex Alpha (Case C–473/98) [2000] ECR I–5681, Swedish legislation prohibiting the sale, transfer or use, for industrial purposes, of chemical products composed wholly or partially of trichloroethylene was held to breach Article 34 (but was justifiable under Article 36). This case is discussed in detail in the next section.

Three cases further illustrate this type of MEQR. In each case the Court held that a complete ban on the use of a product in a Member State breaches Article 34 (even if the product could be lawfully imported and sold) because customers in that state would have little or no interest in buying such a product, if they knew that they could not lawfully use it. Hence the ban creates a potential barrier to trade in that product. Note: in all three cases the legislation, being indistinctly applicable, was at least potentially justifiable using Cassis de Dijon principles.

  • Commission v Portugal (Tinted Film for Car Windows) (Case C–265/06) [2008] ECR I–2245: Portuguese legislation prohibited (with limited exceptions) the ‘affixing of tinted film to the windows of passenger or goods vehicles’.
  • Commission v Italy (Motorcycle Trailers) (Case C–110/05) [2009] ECR I–519: Italian legislation prohibited mopeds and motorcycles from towing trailers.
  • Mickelsson & Roos (Case C–142/05) [2009] ECR I–4273: Swedish legislation prohibited the use of ‘personal watercraft’ – jet-skis – except on water designated as a ‘general navigable waterway’.

14.5 Article 36 and the derogations from Articles 34 and 35

14.5.1 The grounds under Article 36

The grounds listed are exhaustive and may not be added to. In Commission v Ireland (Souvenir Jewellery) (1981), the ECJ stated:

In particular, arguments based on economics have consistently been rejected (Campus Oil Ltd (Case 72/83) [1984] ECR 2727; Evans Medical and Macfarlan Smith (1995)). Other examples of rejected arguments include:

  • consumer protection and/or the fairness of commercial transactions (Souvenir Jewellery (1981));
  • protection of cultural diversity (Leclerc (Case 229/83) [1985] ECR 1).

In that case the ECJ stated:

However, these two grounds have now been recognised as ‘mandatory’ or ‘overriding’ requirements under Cassis de Dijon (1979) principles instead (see section 14.6). The grounds under Article 36 have also been restrictively interpreted because they operate as exceptions from the fundamental freedom of movement. The burden of proving that an Article 36 derogation has been made out rests with the party seeking to rely upon it, usually the national authorities (Denkavit Futtermittel (Case 251/78) [1979] ECR 3369).

In ATRAL (Case C–14/02) [2003] ECR I–4431, the ECJ was asked whether a Member State which claims justification under Article 36 and/or Cassis de Dijon (1979) principles may merely rely on it in the abstract or must specifically demonstrate its genuineness. The ECJ replied (emphasis added):

Public morality

The protection of public morality was successfully invoked in R v Henn and Darby (1979):

The ECJ stated:

The Henn and Darby case was distinguished by the ECJ in another case involving ‘public morality’, Conegate (Case 121/85) [1986] ECR 1007 (below), because that case involved ‘arbitrary’ discrimination against imported goods.

Public policy

Despite great potential width, this ground has rarely been successfully invoked. It does not provide some general fallback provision for states (Commission v Italy (1961)); nor can it be used for purely economic reasons (Commission v Italy (Securities for Imports) (Case 95/81) [1982] ECR 2187). It cannot be used to justify measures that really fall within consumer protection (Kohl v Ringelhan (Case 177/83) [1984] ECR 3651). It is also no justification that the activities with which the law deals are subject to criminal penalties (Prantl (Case 16/83) [1984] ECR 1299). However, the defence was successfully invoked in R v Thompson (Case 7/78) [1979] ECR 2247 (involving UK legislation that imposed an export ban on old coins).

More controversially, the House of Lords used the public policy derogation in R v Chief Constable of Sussex, ex p ITF Ltd [1998] 3 WLR 1260, to justify the Chief Constable’s decision to withdraw police officers from the port of Shoreham in Sussex. This had the effect of allowing animal-rights protestors to blockade the port and prevent companies such as ITF Ltd from exporting live animals to other EU Member States. The Lords decided that this decision was justifiable because the Chief Constable had limited police manpower resources and had to deploy his officers throughout the county, not just in one town.

The public policy derogation was successfully invoked by the Finnish government in Ahokainen & Leppik (Case C–434/04) [2006] ECR I–9171. The case involved a provision of Finnish law, under which a licence was required in order to import drinks containing ethyl alcohol over 80% proof, breach of which was a criminal offence. Two Finnish nationals, Jan-Erik Ahokainen and Mati Leppik, were convicted of ‘smuggling’ – importing without a licence – nearly 10,000 litres of spirits into Finland from Germany and sentenced to prison. On appeal, they argued that the Finnish law amounted to a breach of Article 34. In response, the Finnish government invoked Article 36 – relying on both public health and public policy. The ECJ agreed, stating:

Public security

Again, purely economic reasons will not suffice. However, the presence of an economic justification for national legislation will not be fatal provided that the legislation is also justifiable on public security grounds. The defence was successfully invoked in Campus Oil (1984). Irish legislation restricted the importation of petroleum products, ostensibly to reduce the danger of Ireland becoming over-reliant on imports. The effect of the legislation was that petrol companies operating in Ireland were forced to obtain about 30% of their supplies from Ireland’s only oil refinery, in Cork. The ECJ stated:

Campus Oil (1984) was confirmed in Commission v Greece (Petroleum Stocks) (Case C–398/98) [2001] ECR I–7915. The ECJ acknowledged that ‘the maintenance on national territory of a stock of petroleum products allowing continuity of supplies to be guaranteed constitutes a public security objective’. However, the Court went on to reject the defence on the facts, finding that the Greek legislation was primarily concerned about protecting the economic freedom of oil refineries rather than Greek public security.

Public security was also raised in Richardt and Les Accessoires Scientifiques (Case C–367/89) [1991] ECR I–4621, although this did not involve imports or exports. Instead, it concerned the transit of a machine used in the production of bubble memory circuits from the USA to Russia via Luxembourg. The transit company was prosecuted under Luxembourg law for failing to have acquired the requisite clearance in advance. It challenged the law, and the Luxembourg authorities relied on public security. Bubble memory circuits were a type of computer memory storage that used a thin film of a magnetic material to hold small magnetised areas, known as ‘bubbles’. They were widely used in the 1970s and 1980s but eventually fell into disuse with the advent of hard disks. However, bubble memory circuits found uses through the 1980s in computer systems operating in high vibration or harsh environments. It appears that the Luxembourg authorities were concerned that the machine might have been put to military use, and hence invoked the public security derogation. The ECJ stated:

Protection of the health and life of humans, animals and plants

In Toolex Alpha (2000), where Swedish legislation prohibited the sale, transfer or use, for industrial purposes, of chemical products composed wholly or partially of trichloroethylene, because of a perceived risk of causing cancer, the ECJ stated that ‘the health and life of humans rank foremost among the property or interests protected’ by Article 36. Article 36 has been used to justify many measures, typically prohibitions (or at least restrictions) on the import or sale of various foodstuffs, additives, drugs and chemicals; tests and inspections as to quality, etc.

Sometimes the defence is successful. The Dutch restrictions on the sale of vitamin-enhanced muesli bars in Sandoz (1983), the Swedish ban on trichloroethylene in case Toolex Alpha (2000) and the French ban on ‘Red Bull’ in Commission v France (Case C–24/00) [2004] ECR I–1277 were all upheld. In Scotch Whisky Association (Case C–333/14) [2016] 2 CMLR 27; [2016] 1 WLR 2283, the ECJ held that Scottish legislation, the Alcohol (Minimum Pricing) (Scotland) Act 2012, which permitted the Scottish government to impose a minimum retail price for alcohol, was an MEQR (it ‘significantly restricts the freedom of economic operators to determine their retail selling prices and, consequently, constitutes a serious obstacle to access to the UK market of alcoholic drinks lawfully marketed in [other] Member States’) but one which was justifiable on health grounds. The Court held that the specific aim of the legislation was to raise the price of ‘cheap’ alcoholic drinks. A minimum price for such drinks was ‘capable of reducing the consumption of alcohol, in general, and the hazardous or harmful consumption of alcohol, in particular, given that drinkers whose consumption can be so described purchase, to a great extent, cheap alcoholic drinks’. The Court then left it to the Scottish courts to determine whether the 2012 Act was necessary, given alternative measures such as higher excise duties. Subsequently, in October 2016, the Scottish Court of Session held that the 2012 Act was justified under Article 36, a decision which was later confirmed by the UK Supreme Court (Scotch Whisky Association v Lord Advocate [2017] UKSC 76).

However, sometimes the defence fails. Defences tend be rejected for one of two reasons: (1) lack of evidence of a ‘real’ health risk or (2) failure to comply with the principle of proportionality (see below).

In general, the ECJ is vigilant to prevent Article 36 from being abused, hence the insistence of evidence of a ‘real’ risk. However, if the Member State whose legislation is being contested can point to international scientific research supporting them, then the ECJ is far more likely to accept that there was such a risk. In Green-ham and Abel (2004), the ECJ stated:

Member States may adopt different approaches to the same safety issue. What if there is disputed scientific opinion about whether or not there is a real health risk? In Sandoz (1983), the ECJ stated: ‘Insofar as there are uncertainties at the present state of scientific research it is for the Member States … to decide what degree of protection of the health and life of humans they intend to assure.’ This led the Court to conclude that Dutch rules prohibiting vitamin-enhanced muesli bars were ‘justified on principle’.

In Greenham and Abel (2004), involving French legislation which prohibited the marketing of food and drink to which a substance called Q10 had been added, the ECJ laid down important guidelines on when the human health derogation would be available in such circumstances. The ECJ stated:

The Court repeated this test in Commission v Denmark (Case C–192/01) [2003] ECR I–9693, deciding that Danish rules restricting the marketing of ‘enriched’ foodstuffs (described as any food to which had been added any substance designed to modify its nutritional value, shelf-life, colour, flavour or taste) were justifiable on human health grounds.

In B.S. & C.A. (Case C–663/18) [2021] 2 CMLR 5, the Court held that the cannabidiol (a liquid derived from the cannabis plant) used in the Kanavape e-cigarette produced in the Czech Republic amounted to ‘goods’, as opposed to a ‘narcotic drug’ (refer back to Chapter 10 for a full discussion of this aspect of the case). This meant that the prohibition of the sale of cannabidiol in France under the Public Health Code was an MEQR prima facie prohibited by Article 34, albeit potentially justifiable under Article 36. However, in order to do so, the French authorities would have to demonstrate that cannabidiol posed a ‘genuine threat’ to human health, bearing in mind ‘the results of international scientific research’.

The protection of animal health was invoked in Bluhme (Case C–67/97) [1998] ECR I–8033. Danish law prohibited the keeping of certain bees on the island of Læsø, Denmark. The Court held that this amounted to an MEQR (bees being ‘goods’ for the purposes of Article 34), but it was justifiable under Article 36. The prohibition was designed to ensure the survival of the native Danish brown bee. The Court stated that measures to ‘preserve an indigenous animal population with distinct characteristics contribute to the maintenance of biodiversity by ensuring the survival of the population concerned’.

Protection of national treasures

No case at ECJ level has yet succeeded on these grounds. In any event, it is more likely to apply to measures taken in respect of exports.

Protection of industrial and commercial property

In Belgium v Spain (Rioja Wine Exports) (Case C–388/95) [2000] ECR I–3123, the ECJ upheld provisions of Spanish law requiring Rioja wine intended for export to be bottled in the La Rioja region. Although prima facie in breach of Article 35, the legislation was justified. This was because Rioja wine enjoyed an international reputation for high quality, which might be tarnished if the wine could be transported out of the region in bulk and then bottled elsewhere. The ECJ stated:

14.5.2 The second sentence of Article 36

As well as providing evidence to support the use of one of the six grounds of derogation under Article 36, the Member State in question must also satisfy the Court that no breach of the second sentence of Article 36 has occurred. There are two aspects to this. Any quantitative restriction or MEQR must not:

  • discriminate against imports in an ‘arbitrary’ way; or
  • be a restriction on imports for economic reasons but ‘disguised’ using one of the derogations (usually the ‘health’ derogation).

‘Arbitrary discrimination’

National measures which only apply to imported goods (that is, quantitative restrictions in the strict sense, such as import bans; or distinctly applicable MEQRs) may be justified under Article 36. That is, Member States may discriminate against imported goods. An import ban is, by definition, discriminatory against imports. However, national legislation must not discriminate in an ‘arbitrary’ way. National legislation will be held to be ‘arbitrary’ if there is no objective basis for making the distinction (Commission v France (Case 152/78) [1980] ECR 2299). In R v Henn and Darby (1980) (in the context of the ‘public morality’ derogation) the ECJ said that ‘the true function’ of the second sentence of Article 36 was:

In Henn and Darby (1980), the ECJ regarded s42 of the Customs Consolidation Act 1876 as discriminatory, since it prohibited the importation of material which was ‘indecent or obscene’, whereas domestic (UK-produced) pornography was only illegal under the Obscene Publications Acts 1959 and 1964 if likely to ‘deprave or corrupt’. Clearly, some foreign pornography, unlikely to deprave or corrupt, could still be described as ‘indecent or obscene’. However, the ECJ held that, although discriminatory, the UK law was not arbitrary, and nor was there a disguised restriction on trade. The UK’s anti-pornography laws, taken as a whole, did have as their purpose the prohibition (or at least the restraining) of the manufacture and marketing of articles of an indecent or obscene character.

The 1876 Act had, therefore, been genuinely applied for the protection of public morality, and not for the protection of national products, because there was no lawful trade in such goods in the UK. Henn and Darby was distinguished on that point in Conegate (Case 121/85) [1986] ECR 1007.

A ‘disguised restriction on trade’

A national measure, ostensibly designed to protect human health (for example) may not be protected by Article 36 if, in reality, the measure is ‘a disguised restriction on trade’. A good example is provided by Commission v UK (French Turkeys) (Case 40/82) [1982] ECR 2793. In 1981 the UK had banned poultry imports, ostensibly because of fears about a health risk. On closer examination, it transpired that the import ban had been imposed for economic reasons. The ECJ stated:

14.5.3 Article 36 and harmonising Directives

Where harmonising Directives in a particular subject (typically human or animal health) have been adopted, Member States may not unilaterally adopt, on their own authority, corrective or protective measures designed to obviate any breach by another state of EU law. Article 36 will not be available. In Tedeschi v Denkavit (Case 5/77) [1977] ECR 1555, the ECJ stated:

The ECJ confirmed this ruling in two cases involving measures taken allegedly in the interests of animal health: Hedley Lomas (Ireland) Ltd (Case C–5/94) [1996] ECR I–2553 (conditions in Spanish slaughterhouses) and Compassion in World Farming Ltd (Case C–1/96) [1998] ECR I–1251 (live animal exports in crates). In both cases the ECJ held that the measures could not be justified using Article 36 because harmonising Directives provided EU-wide protection.

14.6 The effects of the Cassis de Dijon principle

14.6.1 Introduction

By 1979, Article 36 was already 22 years old and, never having been updated, was ‘stuck’ in 1957. By the late 1970s several Member States (in particular Germany and the UK) had enacted quite sophisticated consumer protection legislation. This legislation was quite capable of restricting the free movement of goods (especially given the wide scope awarded to Article 34 by Dassonville (1974)) but was incapable of being justified under Article 36 unless it could be brought under the ‘protection of human health’ heading. As only the Member States’ governments had the power to amend the EC Treaty (and even then they must be unanimous) the wait for Article 36 to be updated was likely to be a very long wait indeed. In 1979, therefore, in Cassis de Dijon (Case 120/78) [1979] 649, the ECJ took decisive action. As the ECJ had no power to amend Article 36, it created a parallel set of derogations that Member States could plead as an alternative to Article 36. This is sometimes referred to as the ‘rule of reason’.

The ECJ held:

‘Rule of reason’: key points

  • In Cassis de Dijon (1979), the ECJ described the derogations that Member States can plead as ‘mandatory requirements’. It has used other, synonymous, expressions too, such as ‘objectives of general interest’ (ADBHU (Case 240/83) [1985] ECR 531), ‘overriding interests’, ‘overriding requirements’ (Familiapress (Case C–368/95) [1997] ECR I–3689) or ‘overriding reason in the general interest’ (Decker (Case C–120/95) [1998] ECR I–1831).
  • In Cassis de Dijon (1979), the ECJ gave four examples of ‘mandatory requirements’ but prefaced its list with the words ‘in particular’ – thus, there is no limit to the number of mandatory requirements that can be created. This contrasts with Article 36, which is a closed list.
  • In subsequent case law the ECJ has added to the list, which now includes the improvement of working conditions (1981), the protection of culture (1985), the protection of the environment (1985), the diversity of the press (1997), the maintenance of social security systems (1998), road safety (2000), the protection of ‘fundamental freedoms’ such as the freedom of expression (2003), the protection of children (2008) and the ‘fight against crime’ (2008).

14.6.2 The ‘mandatory requirements’

The protection of public health (1979)

Although generally dealt with under Article 36, in one case at least, public health was dealt with under Cassis de Dijon (1979) principles (Gilli and Andres (1981)).

The fairness of commercial transactions (1979)

This was successfully argued in Oosthoek’s (Case 286/81) [1982] ECR 4575. The Dutch government argued that legislation prohibiting the offering of free gifts as a means of sales promotion was justifiable on grounds of fair trading (and consumer protection). The ECJ agreed. The ECJ held that it was ‘undeniable’ that offering free gifts ‘may mislead customers as to the real prices of certain products and distort the conditions on which genuine competition is based’. In IDG (Case 6/81) [1982] ECR 707, the ECJ agreed that Dutch legislation prohibiting the marketing of identical products (known as ‘passing off’ in English law) was justifiable on grounds of fair trading.

Fair trading was claimed, but rejected, in another ‘passing off’ case, that of Prantl (Case 16/83) [1984] ECR 1299.

The Court stated:

The defence of the consumer (1979)

The defence has been successfully raised in some cases, such as Robertson and Others (Case 220/81) [1983] ECR 2349, concerning Belgian legislation requiring the hallmarking of silver-plated goods. Another example of the defence being successfully raised is A-Punkt Schmuckhandels v Schmidt (Case C–441/04) [2006] ECR I–2093. The case involved an alleged breach by Claudia Schmidt, a jeweller, of the Austrian Trade and Commercial Regulations 1994, which prohibited the sale of silver jewel-lery to customers in their own homes. Although the Court held that the 1994 Regulations probably amounted to a ‘selling arrangement’, in which case Article 34 did not apply (see section 14.8), the ECJ also considered the possibility that the Regulations breached Article 34 and, if so, whether they could be justified using the consumer protection mandatory requirement. On that point, the Court stated:

However, the defence has, in the majority of cases, proven to be unsuccessful. It is clear that the ECJ attributes a fair amount of intelligence and sophistication to consumers. Indeed, in Estée Lauder v Lancaster (2000) the ECJ held: ‘It is necessary to take into account the presumed expectations of an average consumer who is reasonably well informed and reasonably observant and circumspect.’

In Clinique Laboratories (1994) and Mars (1995) the ECJ rejected the ‘consumer protection’ defence, which had been advanced by the German government in order to restrict the marketing of French goods (cosmetics and ice-cream bars, respectively). In Mars, the German authorities had objected to the sale of ice-cream bars bearing a ‘+10%’ logo which covered substantially more than 10% of the surface area of the wrapping. The ECJ stated:

A consumer protection defence was also rejected in Clinique Laboratories (1994):

Often the defence will be rejected because the contested national legislation went beyond what was ‘necessary’ to protect consumers. Thus, in Walter Rau v De Smedt (1982), the ECJ held that the Belgian legislation requiring margarine to be packed in cubes was prohibited by Article 34, despite alleged consumer protection reasons, because Belgian consumers could have been adequately protected by less drastic means, e.g. clear labelling.

In Alfa Vita Vassilopoulos (Cases C–158 and 159/04) [2006] ECR I–8135, the ECJ emphasised that the mandatory requirement involved protection of the consumer, not the promotion of quality. The Court therefore held that the defence was unavailable to justify a provision in Greek law designed to help customers distinguish between fresh bread and ‘bake-off’ bread – meaning, bread which had previously been frozen – the former, apparently, being superior quality to the latter. The Court stated:

The improvement of working conditions (1981)

The improvement of working conditions was added as a mandatory requirement in 1981 in Oebel (Case 155/80) [1981] ECR 1993, concerning German rules preventing bakeries from operating before 4 a.m. Although this was capable of restricting trade (in that it prevented German bakeries from producing bread early enough for export to pre-breakfast markets just over the border in the neighbouring states of Luxembourg and Belgium), the ECJ held that the restriction was justified. The Court stated: ‘It cannot be disputed that the prohibition in the bread and confectionery industry on working before 4 am in itself constitutes a legitimate element of economic and social policy, consistent with the objectives of public interest pursued by the Treaty.’

The protection of the environment (1985)

The European Commission first mentioned this in a Practice Note in 1980. The Commission, ostensibly summarising the list of ‘mandatory requirements’ at that date, listed ‘public health, protection of consumers or the environment, the fairness of commercial transactions, etc.’ However, environmental protection was only recognised as a mandatory requirement by the ECJ in 1985 in ADBHU (Case 240/83) [1985] ECR 531, involving French legislation restricting the movement of waste oil. The ECJ stated:

Insofar as such measures … have a restrictive effect on the freedom of trade … they must nevertheless neither be discriminatory nor go beyond the inevitable restrictions which are justified by the pursuit of the objective of environmental protection, which is in the general interest.

The best-known case is Commission v Denmark (the Danish Bottles case) (Case 302/86) [1988] ECR 4607, where the ECJ stated:

Environmental protection was also accepted as a defence in Commission v Belgium (Walloon Waste) (Case C–2/90) [1992] ECR I–4431, involving Belgian legislation restricting the importation of waste (for recycling purposes) and Aher-waggon (Case C–389/96) [1998] ECR I–4473, involving German legislation imposing strict noise levels on aircraft engines. The ECJ rejected an environmental protection defence in Dusseldorp (Case C–203/96) [1998] ECR I–4075, involving Dutch legislation restricting the exportation of waste oil filters, but only because the legislation was ‘primarily’ economic in nature.

Two more cases in which environmental protection was accepted as a defence, at least in principle, are Commission v Germany (Waste Packaging) (Case C–463/01) [2004] ECR I–11705 and Radlberger & Spitz (Case C–309/02) [2004] ECR I–11763. Both cases involved German legislation providing for a compulsory deposit-and-return scheme for waste packaging, which meant that manufacturers were obliged to charge a deposit for, and accept the return of, waste packaging. Although the Court accepted that the German rules were capable of hindering trade, they were justifiable. The Court stated:

In Mickelsson & Roos (Case C–142/05) [2009] ECR I–4273, Swedish legislation prohibiting the use of ‘personal watercraft’ – jet-skis – except on water designated as a ‘general navigable waterway’ was held to breach Article 34. However, it was justifiable on environmental protection grounds. The Court held:

In Ålands Vindkraft (Case C–573/12) [2015] 1 CMLR 10 and Essent Belgium (Case C–492/14) [2017] 1 CMLR 39, the Court examined whether national legislation encouraging the use of renewable energy was capable of justification under Cassis de Dijon in the event that the legislation was found to impose an obstacle to free trade. In both cases the answer was ‘yes’, at least in principle; although only the legislation in the former case passed the proportionality test and was held to be justified.

The protection of culture (1985)

This was added to the list in 1985 by Cinéthèque (Cases 60 and 61/84) [1985] ECR 2065, concerning French legislation temporarily restricting the availability of films on video, in order to encourage cinema attendance instead. The ECJ held that this policy was justifiable. (It is perhaps not insignificant that this case was heard at a time when cinema attendances were declining across Europe, largely because of the introduction of home video recorders in the early 1980s.)

In Torfaen Borough Council v B&Q plc (Case 145/88) [1989] ECR 3851, which concerned the legality of British Sunday trading legislation (the Shops Act 1950), the ECJ first coined the phrase ‘sociocultural characteristics’ to describe this mandatory requirement. The ECJ held that the prohibition on Sunday trading, although capable of hindering trade, was justifiable. (Note: Sunday trading legislation has subsequently been re-classified as a ‘selling arrangement’, which is exempt from EU law – see section 14.8.2.)

The protection of books as ‘cultural objects’ was recognised as an overriding interest in Fachverband der Buch- und Medienwirtschaft (Case C–531/07) [2009] ECR I–3717. Austrian legislation stipulated a minimum selling price for books. When challenged to justify this legislation, the Austrian government argued that, in the absence of a pricing system, there would be a drop in prices, which would cause a drop in profits, as a result of which it would become impossible to finance the production and marketing of more demanding but economically less attractive works. Moreover, small booksellers which normally offer a wider choice of specialist books would be driven out of the market by larger booksellers which sell primarily more commercial books. The ECJ accepted that books were ‘cultural objects’ deserving of protection, although the actual legislation was deemed to go beyond what was necessary to achieve that objective (in other words, it failed the proportionality test).

The diversity of the press (1997)

This was added to the list in 1997 by Familiapress (1997), concerning a prohibition in Austrian legislation on newspapers offering cash prizes to competition winners. The ECJ decided that the Austrian legislation was justifiable in the interest of helping smaller publishers to survive against fierce competition from larger publishers (who had the potential to offer bigger prizes), thereby promoting a diverse newspaper industry. The ECJ stated: ‘Maintenance of press diversity may constitute an overriding requirement justifying a restriction on free movement of goods. Such diversity helps to safeguard freedom of expression.’

The maintenance of social security systems (1998)

This was added to the list in 1998, in Decker (Case C–120/95) [1998] ECR I–1831. D, a Luxembourg national, used a prescription obtained in Luxembourg to purchase a pair of spectacles from an optician in Belgium. When he tried to reclaim the cost from the Luxembourg social security authorities, this was refused as he had not sought prior authorisation. This rule only applied where spectacles were obtained abroad. D argued that this was contrary to Article 34. The ECJ held that the Luxembourg rules were prima facie in breach of Article 34, because they encouraged Luxembourg nationals to purchase or have their spectacles assembled in Luxembourg rather than another Member State, such as Belgium. However, the rules were justifiable under Cassis de Dijon (1979) principles. The Court stated that ‘the risk of seriously undermining the financial balance of the social security system may constitute an overriding reason in the general interest’.

Road safety (2000)

In Snellers Autos (Case C–314/98) [2000] ECR I–8633, the ECJ confirmed that road safety could be added to the list of mandatory requirements. This was confirmed in

Commission v Finland (Case C–54/05) [2007] ECR I–2473, involving Finnish legislation under which a ‘transfer licence’ was required before cars registered in other Member States could be imported into Finland. The licence was not automatic but was regarded in practice as a formality. The European Commission alleged a breach of Article 34, and in response the Finnish government argued that the licensing requirement was justified on the grounds of promoting road safety (licences could be refused to cars not deemed to be roadworthy). The ECJ agreed (in principle, at least), stating: ‘It is not in dispute that road safety does constitute an overriding reason in the public interest capable of justifying a hindrance to the free movement of goods.’

In Commission v Italy (Motorcycle Trailers) (Case C–110/05) [2009] ECR I–519, involving Italian legislation prohibiting mopeds and motorcycles from towing trailers, the ECJ held that the legislation was capable of hindering trade in such trailers but was justified on grounds of road safety. The Court stated:

In Commission v Poland (Right-Hand Drive Cars) (Case C–639/11) [2014] 3 CMLR 26, the Court had to decide whether Polish legislation requiring drivers of imported right-hand drive cars to convert their cars to left-hand drive was compatible with EU law on the free movement of goods.

The protection of fundamental rights (2003)

In Schmidberger v Austria (Case C–112/00) [2003] ECR I–5659, the ECJ stated that the protection of fundamental rights ‘is a legitimate interest which, in principle, justifies a restriction of … the free movement of goods’. The case involved a decision made by the Austrian authorities to allow a road to be closed for the purposes of a demonstration. The Court acknowledged that the authorities had been placed in a difficult position, having to balance the right of transport companies to enjoy the free movement of goods under Articles 34 and 35, on one hand, against the rights of the demonstrators to enjoy the freedoms of expression and assembly, under Articles 10 and 11 of the European Convention on Human Rights, on the other. The ECJ held that no breach of EU law had occurred.

The Court stated:

Protection of children (2008)

The case of Dynamic Medien v Avides Media (Case C–244/06) [2008] ECR I–505, involved a system under German law of compulsory classification and labelling of DVDs and videos in terms of their suitability for viewing by children. The Court accepted that, although capable of hindering trade, the German rules were justifiable. However, it is not entirely clear whether the Court dealt with the justification as an example of public policy under Article 36, or as a new mandatory requirement. At one point, the Court observed that the European Commission, along with the British, German and Irish governments, all agreed that ‘the rules at issue … are justified in so far as they are designed to protect young people. That objective is linked in particular to public morality and public policy, which are grounds of justification recognised in [Article 36].’ However, the Court then went on to state that ‘the protection of the child is a legitimate interest which, in principle, justifies a restriction on a fundamental freedom guaranteed by the EC Treaty, such as the free movement of goods’, which seems to suggest a new mandatory requirement.

Whichever is the true position, the Court accepted that the German rules ensured ‘that young people are able to develop their sense of personal responsibility and their sociability. Furthermore, the protection of young people is an objective which is closely related to ensuring respect for human dignity’. The Court accepted that the German rules were ‘designed to protect children against information and materials injurious to their well-being’ and that, in such a case, in view of the divergence of ‘moral or cultural views, Member States must be recognised as having a definite margin of discretion’.

The ‘fight against crime’ (2008)

The ‘fight against crime’ was added to the list in Commission v Portugal (Tinted Film for Car Windows) (Case C–265/06) [2008] ECR I–2245. The Portuguese government accepted that legislation banning the affixing of tinted film to the windows of cars and other vehicles was capable of hindering trade but argued that it was justified on grounds of public safety and/or road safety. The ban was:

intended to enable the competent authorities to make a rapid external inspection of the interior of motor vehicles without the need to immobilise them, first, in order to ensure that the vehicle’s occupants are wearing seat belts and, second, to identify potential criminals for the purpose of combating crime.

The ECJ accepted these arguments, at least in principle, stating: ‘The fight against crime and ensuring road safety may constitute overriding reasons in the public interest capable of justifying a hindrance to the free movement of goods.’

14.6.3 Application to ‘indistinctly applicable’ measures only?

The traditional position of the ECJ has been that the ‘mandatory requirements’ are only available when national legislation is ‘indistinctly applicable’. In Gilli and Andres (1981), for example, the ECJ stated:

The ECJ confirmed this in Commission v Ireland (Souvenir Jewellery) (1981), rejecting a consumer protection defence on the basis that the contested Irish legislation only applied to imported jewellery.

This means that Cassis de Dijon (1979) should, logically, never be available in Article 35 cases, because the ECJ has held that national legislation only breaches Article 35 if it is ‘distinctly applicable’ (Bouhelier (Case 53/76) [1977] ECR 197 or if it has as its ‘specific object or effect the restriction of patterns of exports’ (PB Groenveld (Case 15/79) [1979] ECR 3409, discussed below)). However, in two of the cases discussed above the ECJ accepted that Cassis de Dijon may apply in the context of Article 35 (Oebel (1981) and Dusseldorp (1998)).

This area of doubt has now been clarified by the ECJ. Gysbrechts & Santurel Inter (Case C–205/07) [2008] ECR I–9947 involved Belgian legislation under which it was illegal for Belgian retailers to require any payment from consumers within the seven-day ‘cooling off’ period allowed for distance-selling contracts. The ECJ accepted that the Belgian legislation had the effect of potentially restricting exports, as it deprived Belgian firms of an effective form of protection against defaulting consumers based in other Member States (given the difficulty of bringing legal proceedings in other jurisdictions). However, the Belgian legislation was held to be justifiable on consumer protection grounds, under Cassis de Dijon principles.

The Court stated:

14.7 General rules concerning derogations: proportionality and mutual recognition

14.7.1 Proportionality

The ECJ has consistently held that the purpose of Article 36 is to allow certain national laws and rules to derogate from the free movement provisions only to the extent to which they are ‘justified’ in order to achieve the objectives in the Article. A measure may be justified provided it does what is necessary to achieve the objectives in the first sentence of Article 36, and further that it does no more than necessary. If there are other methods capable of achieving that objective which are less restrictive of intra-community trade, then they should be used instead. In De Peijper (Case 104/75) [1976] ECR 613, for example, the ECJ said:

Many of the cases where Article 36 is invoked involve national measures introduced to protect against alleged risks to human health. Often, where a particular ingredient or additive has been restricted or prohibited altogether, the ECJ will pose the (rhetorical) question whether human health could still be protected by the simple expedient of requiring manufacturers of products to label the ingredients in their products clearly. Thus, in Commission v France (Case C–24/00) [2004] ECR I–1277, for example, the ECJ stated:

The same principles apply to attempts by Member States to justify legislation under Cassis de Dijon (1979) principles. The best example is perhaps Walter Rau v De Smedt (1982), discussed above, where the ECJ stated: ‘if a Member State has a choice between various measures to attain the same objective it should choose the means which least restricts the free movement of goods’.

In Commission v France (2004), the ECJ summarised the proportionality principle as follows:

14.7.2 Mutual recognition

As well as creating the ‘mandatory requirements’ in Cassis de Dijon (1979), the ECJ went on to establish a presumption that, once goods have been ‘lawfully produced and marketed in one of the Member States’, they may be imported into any other state. This has become known as the ‘mutual recognition’ principle. The presumption may only be rebutted by evidence that the goods in question pose a threat to one of the heads or Article 36 or one of the ‘mandatory requirements’. The net result is to place the burden of proof on the authorities of the Member States seeking to justify their domestic legislation. The best way to rebut the presumption is to identify specific national characteristics (usually involving dietary habits) which would justify different national legislative provisions. In Muller (1986), dealing with E475, a baking ingredient allowed in Germany but not in France, the ECJ stated:

Similar principles are seen in these cases:

  • Bellon (Case C–42/90) [1990] ECR I–4863: French legislation banned certain Italian pastries because they contained sorbic acid. The ECJ held that the legislation was justifiable on health grounds, but only after taking national dietary habits into account.
  • Aher-Waggon (1998): German legislation set aircraft noise emission levels, which were more restrictive than in other states. The ECJ stated that these rules were justified because Germany ‘is a very densely populated State’ and therefore ‘attaches special importance to ensuring that its population is protected from excessive noise emissions’.
  • Greenham and Abel (2004): dealing with French legislation that banned, inter alia, an ingredient in fruit drinks called coenzyme Q10, the ECJ held that the ban would be permissible on health grounds only if the French authorities could show ‘in the light of national nutritional habits and in the light of the results of international scientific research, that their rules are necessary’.

In practice, rebutting the presumption will not be easy to do. In Mars (1995) and Clinique Laboratories (1994), German consumer protection arguments were rejected. The ECJ held in each case that the German legislation was in breach of Article 34. As the goods concerned (ice creams and cosmetics, respectively) were lawfully manufactured and marketed in France, the mutual recognition principle applied, and the German government had failed to show why it needed stricter laws than France. In Commission v Germany (the Beer Purity case) (1987), concerning German legislation on the additives permitted in beer, the German government tried to argue that it needed stricter rules on beer purity than other Member States because German nationals tended to drink more beer than nationals of other Member States. However, although not disputing that assertion, the ECJ did not accept that it justified stringent rules which made it practically impossible for French-made beer (containing banned additives) to be sold in Germany. The Court stated:

14.8 The divisions in Keck and Mithouard: ‘Selling arrangements’

14.8.1 Introduction

It has been noted already that Article 34 was widely defined in Dassonville (1974). However, in 1993, the ECJ acknowledged that the Dassonville ‘formula’ was so wide that it was leading importers and retailers to challenge a whole range of national laws whose likely impact on the free movement of goods was, at most, marginal. In the landmark ruling in Keck and Mithouard (1993), the ECJ announced that there was a distinction to be drawn between:

  • Product requirements: these are laws regulating the goods themselves, which are still governed by Article 34 and the Dassonville formula and are prohibited. Such rules are prima facie contrary to EU law and require justification, under either Article 36 or Cassis de Dijon (1979) principles.
  • Selling arrangements: these are laws concerning not the goods themselves, but rather how, when and where they are marketed. These rules fall outside of the scope of Article 34 altogether and hence do not require justification. In Keck, the ECJ said that these rules were prima facie lawful, although they were subject to two pre-conditions.

Strictly speaking, the decision in Keck was not entirely unprecedented. In a handful of earlier cases, the ECJ had held that national rules governing ‘selling arrangements’ are exempt from Article 34. In the earliest cases, Blesgen (Case 75/81) [1982] ECR 1211, involving Belgian legislation prohibiting the public consumption of strong alcohol, the ECJ stated: ‘Such a legislative measure has no connection with the importation of the products and for that reason is not of such a nature as to impede trade between Member States.’ Then, in Quietlynn v Southend BC (Case C–23/89) [1990] ECR I–3059, which involved UK law regulating licences for sex shops, the ECJ again ruled that Article 34 did not apply.

The logical conclusion of these developments occurred in Keck, which involved French legislation preventing the resale of goods at a loss, when the ECJ made a decisive statement on the legality of ‘selling arrangements’. The ECJ began by conceding that the French legislation might have an effect on inter-state trade. It considered that such legislation ‘may, admittedly, restrict the volume of sales of products from other Member States, insofar as it deprives traders of a method of sales promotion’. However, the ECJ then observed that ‘national legislation imposing a general prohibition on resale at a loss is not designed to regulate trade in goods between Member States’. The ECJ concluded that, ‘in view of the increasing tendency of traders to seek to avoid non-protectionist national laws’ by relying on Article 34, it was necessary to review its position. It went on:

You will note that the Court overruled any previous conflicting judgments but failed to identify what they are! Since 1993, therefore, there has been considerable discussion about which cases were overruled by Keck and which were not.

14.8.2 Examples of selling arrangements

Sunday trading legislation

Prior to Keck, the ECJ’s policy was that national legislation regulating the freedom of retailers to trade on Sunday infringed Article 34 because it had the potential to hinder trade. One such example was the UK’s Shops Act 1950 (subsequently amended by the Sunday Trading Act 1994). In Torfaen BC v B&Q plc (1989) (discussed above in the context of the Cassis de Dijon principle) and Stoke-on-Trent CC v B&Q plc (Case C–169/91) [1992] ECR I–6635, the ECJ held that Article 34 was infringed. However, after Keck, the ECJ held that Article 34 did not apply to such legislation at all. This new approach was set down in a series of cases concerning Italian Sunday trading legislation (Punto Casa and PPV (Cases C–69 and 258/93) [1994] ECR I–2355, and Semeraro Casa Uno and Others (Cases C–418 to 421/93) [1996] ECR I–2975). In Pelckmans Turnhout (Case C–483/12) [2014] 3 CMLR 49, involving Belgian legislation which required retailers to close for one day per week, the ECJ confirmed that such legislation was a ‘selling arrangement’ (following Punto Casa and PPV and Semeraro Casa Uno and Others).

Opening hours

In Tankstation t’Heukske and B E Boermans (Cases C–401 and 402/92) [1994] ECR I–2199, concerning Dutch rules on the opening hours of petrol stations, the ECJ stated the rules laid down in Keck were satisfied. National licensing legislation (such as that in England and Wales, whereby the normal closing hours for public houses is 11 p.m.) would also come within this heading.

Sale of goods only through specific outlets

With the benefit of hindsight, the case of Quietlynn (1990) (discussed above) falls into this category. Another example is Commission v Greece (Processed Milk) (Case C–391/92) [1995] ECR I–1621, where Greek legislation prohibited the sale of processed milk for infants except from pharmacies. The ECJ stated that the Greek rules concerned selling arrangements, and not goods themselves, and were therefore exempt from Article 34.

Restrictions on certain forms of sale promotions

In a pre-Keck case, Oosthoek’s (Case 286/81) [1982] ECR 4575, the ECJ ruled that Dutch legislation prohibiting the offering of free gifts as a sales promotion strategy was prohibited by Article 34. Similarly, in Buet and EBS (Case 382/87) [1989] ECR 1235, the ECJ ruled that very similar French legislation prohibiting door-to-door canvassing of educational materials could hinder trade. However, it can now be argued that the Dutch and French laws in question were ‘selling arrangements’.

For example, in Burmanjer & Others (Case C–20/03) [2005] ECR I–4133, three Dutch nationals were accused of breaching Belgian legislation prohibiting the sale of subscriptions to periodicals (such as magazines) on the street without prior authorisation. In their defence they argued that the Belgian law was contrary to Article 34. The ECJ, however, held that the legislation was a selling arrangement and therefore exempt from Article 34. This was because the legislation did not ban the product, or even the sale of it, it simply banned one method of sale. In the more recent case of DocMorris (Case C–190/20) (2021), unreported, the Court accepted that German legislation prohibiting pharmaceutical companies from offering cash and/or promotional gifts as an incentive to prospective purchasers of prescription medicinal products was a ‘selling arrangement’ and therefore exempt from Article 34.

Restrictions on advertising

In GB-INNO-BM (Case 362/88) [1990] ECR I–667, the ECJ was asked to rule on the compatibility of Article 34 with Luxembourg legislation preventing advertising campaigns that made reference to the pre-sale price of a product. The ECJ held that the legislation was prohibited by Article 34 because it was capable of hindering trade. Similarly, in Aragonesa and Publivia (1991), the ECJ held that national legislation prohibiting the advertising of alcohol over a certain strength in mass media, on streets and highways, in cinemas and on public transport fell within the Dassonville formula.

After Keck, it has become clear that national rules imposing partial restrictions on advertising, at least, are not caught by Article 34. In Hünermund (Case C–292/92) [1993] ECR I–6787, concerning German legislation preventing pharmacists from advertising on the radio, on TV or at the cinema, the ECJ stated that the conditions in Keck were satisfied and the German legislation was therefore exempt from Article 34. The ECJ reached the same conclusion in Leclerc-Siplec (Case C–412/93) [1995] ECR I–179, concerning French legislation prohibiting TV advertising by the distribution sector (which included petrol companies) and in PRO Sieben Media (Case C–6/98) [1999] ECR I–7599, concerning German legislation on the division of time between programmes and advertising on German television. As will be seen below, however, national rules imposing total prohibitions on advertising are not ‘selling arrangements’ and continue to be caught by Article 34.

Requirements as to how goods are presented in shops

It is well established that national rules imposing packaging requirements on products constitute measures equivalent to quantitative restrictions (Walter Rau v De Smedt (1982) being the classic example). However, in Morellato (Case C–416/00) [2003] ECR I–9343, the ECJ decided that a requirement under Italian law that bread had to be sold wrapped constituted a selling arrangement. The crucial difference between Walter Rau (Article 34 and Dassonville applies) and Morellato (2003) (Keck applies) seems to be that in the former case the onus is placed on the manufacturer, while in the latter situation the onus is placed on the retailer. In Morellato, the ECJ stated that the bread could be imported unwrapped and then sold wrapped, with the retailer carrying out the wrapping task. In this way, the wrapping could be regarded as ‘a simple transformation process’ and was therefore incapable of restricting the free movement of goods.

Price controls

Prior to Keck, it had been held that price controls (that is, national rules imposing either maximum or minimum prices on certain products, or rules preventing retailers from selling goods at a loss) could infringe Article 34. However, price controls have now been reclassified as selling arrangements. Keck itself is such a case, as is Belgapom (Case C–63/94) [1995] ECR I–2467, where Keck was applied. A more recent example is Colruyt (Case C–221/15) (2016) (unreported), in which the Belgian supermarket chain Colruyt was prosecuted and fined for selling tobacco products at below cost price, a practice prohibited under Belgian legislation. The company challenged the fine, contending that the Belgian law was precluded by Article 34, but the ECJ held that the Belgian legislation was a ‘selling arrangement’ exempted from Article 34.

Restrictions on the display of goods

Although there is no ECJ case law on this point, in Philip Morris v Norway (Case E–16/10) [2012] 1 CMLR 24, the EFTA Court held that Norwegian law which banned the visible display of tobacco products and smoking devices in Norway was, in principle at least, a ‘selling arrangement’, provided that both Keck conditions were met (which was a decision for the national court). If one or both conditions were not met, then the Norwegian law would be an MEQR, albeit one justifiable on health grounds, subject to the proportionality test. (Note: the EFTA Court in Luxembourg hears disputes involving the interpretation of the European Economic Area Agreement in cases originating in Iceland, Liechtenstein and Norway.)

14.8.3 Failure to satisfy the conditions in Keck and Mithouard

If a provision of national legislation appears to be a ‘selling arrangement’ but actually fails one of the conditions in Keck, then the legislation falls to be dealt with under Article 34 as a measure having an effect equivalent to a quantitative restriction.

The first condition

The first condition is that the national rules alleged to constitute a selling arrangement ‘apply to all relevant traders operating within the national territory’.

The second condition

The second condition is that those provisions must ‘affect in the same manner, in law and in fact, the marketing of domestic products and of those from other Member States’. There have been several cases examining this condition. In Ortscheit (Case C–320/93) [1994] ECR I–5243, the ECJ ruled, distinguishing Keck, that a prohibition on advertising of all foreign medicinal products was caught by Article 34 (although justifiable on health grounds under Article 36). The second condition was not satisfied.

Similarly, in De Agostini and TV-Shop (Cases C–34 to 36/95) [1997] ECR I–3843, concerning Swedish legislation imposing an outright ban on advertising aimed at minors (children less than 12 years of age), the ECJ thought that the legislation might not satisfy the second Keck condition. It actually left the final decision on this issue to the Swedish court which had referred the case to the ECJ in the first place. In Gourmet International Products (Case C–405/98) [2001] ECR I–1795, involving an absolute prohibition in Swedish legislation on the advertising of alcohol, the ECJ decided that the second condition was definitely not satisfied.

Thus, the Swedish legislation constituted an indistinctly applicable MEQR, prima facie prohibited by Article 34 although theoretically justifiable under Article 36 on health grounds.

The ECJ stated:

Gourmet International Products was followed in Douwe Egberts (Case C–239/02) [2004] ECR I–7007. Belgian legislation prohibited any references on the labelling of food to the word ‘slimming’. Douwe Egberts, a Dutch coffee producer, alleged that various statements, such as ‘the absolute breakthrough in weight control’ on the labels of ‘DynaSvelte Café’, a rival brand, infringed this legislation. One issue for the ECJ was whether the Belgian legislation was an MEQR, or whether it was a selling arrangement. The Court confirmed that a total advertising ban could not be categorised as a selling arrangement. The Court stated: ‘An absolute prohibition of advertising the characteristics of a product is liable to impede access to the market by products from other Member States more than it impedes access by domestic products, with which consumers are more familiar.’ Having categorised the Belgian legislation as an MEQR, the Court went on to consider whether it could be justified, on the grounds of the protection of human health, but rejected that on proportionality grounds.

Similarly, in Deutscher Apothekerverband (Case C–322/01) [2003] ECR I–14887, the ECJ again held that the second condition had not been satisfied.

Deutscher Apothekerverband was followed in Ker-Optika (Case C–108/09) [2010] ECR I–12213, involving Hungarian legislation which stipulated that contact lenses could only be sold in specialist medical accessory shops or by home delivery. The effect of this was to prohibit the online sale, in Hungary, of contact lenses. The ECJ held that the Hungarian legislation was not a selling arrangement, because the second Keck condition was not satisfied. The ban on Internet sales had a greater impact, in fact, on traders based in other Member States than it did on traders based in Hungary itself. The Court stated (emphasis added):

the prohibition on selling contact lenses by mail order deprives traders from other Member States of a particularly effective means of selling those products and thus significantly impedes access of those traders to the market of the Member State concerned.

The legislation was therefore an MEQR, prohibited by Article 34, subject to justification under Article 36 on health grounds. Deutscher Apothekerverband was again followed in DPV (Case C–148/15) [2017] 2 CMLR 1. This case involved a provision of German legislation which imposed a series of uniform prices for pharmaceutical products. This was held to be an MEQR, and not a ‘selling arrangement’, because the legislation did not affect the sale of all pharmaceutical products equally. The Court held that mail-order pharmacies based in other Member States were heavily reliant on ‘price competition’ as a means of gaining access to the German market, whereas pharmacies based in Germany had an alternative business model, i.e. selling medicines over the counter from a physical pharmacy. The latter also had an advantage when it came to dispensing medicines in an emergency.

14.9 Article 35 and exports

There are some fundamental differences between the operation of Article 35 and that of Article 34. The case law relating to Article 35 draws an important distinction between distinctly and indistinctly applicable national rules.

14.9.1 ‘Distinctly applicable’ rules

Measures which clearly discriminate against exports will usually be found to be in breach of Article 35. The most obvious examples are export bans, as in:

  • R v Thompson (1979) – UK law prohibited the exportation of old coins. Held to be a prima facie breach of Article 35 (although justifiable under Article 36 on public policy grounds).
  • Hedley Lomas (Ireland) Ltd (1996) – the Ministry of Agriculture, Fisheries and Food (MAFF) banned live animal exports to Spain. This was held to breach Article 35 (and the ECJ rejected a defence of protection of animal health under Article 36 for lack of evidence).
  • Dusseldorp (1998) – prohibition in Dutch law on the exportation of waste oil filters.

This was held to breach Article 35 (and was not justifiable under Cassis de Dijon principles on environmental protection grounds, because the ban was primarily motivated by economic considerations).

The same considerations apply to distinctly applicable MEQRs. Thus, in Jongeneel Kaas (Case 237/82) [1984] ECR 483, Dutch law requiring cheese exporters to be in possession of an export licence was held to be in breach of Article 35. Similarly, in Bouhelier (Case 53/76) [1977] ECR 197, concerning French rules requiring watch-makers to be licensed for export, ostensibly to ensure quality standards were maintained, the ECJ said that Article 35 had been infringed. The Court drew particular attention to the fact that the rules only applied ‘to products intended for export’ and were not imposed ‘on products marketed within the Member State’. This leads to ‘arbitrary discrimination between the two types of products which constitutes an obstacle to intra-Union trade’.

Similarly, in Belgium v Spain (Rioja Wine Exports) (Case C–388/95) [2000] ECR I–3123, Spanish rules on the transportation of Rioja wine drew a distinction between wine intended for export (this wine had to be transported in bottles) and wine intended for distribution to consumers in Spain (this could be transported in bulk). The ECJ held that this rule constituted a breach of Article 35, although justifiable on grounds of protection of commercial property under Article 36.

In Jersey Potatoes (Case C–293/02) [2005] ECR I–9543, the ECJ again held that national legislation which only applied to goods intended for export was capable of breaching Article 35. Under the Jersey Potato Export Marketing Scheme Act 2001, potato producers in Jersey were prohibited from exporting potatoes to mainland UK unless they were registered with the Jersey Potato Export Marketing Board. Failure to comply could lead to fines or imprisonment or both. The Jersey Produce Marketing Organisation contended that the provisions in the 2001 Act constituted a breach of Article 35. The Court agreed, stating: ‘such legislation is, by its very nature, likely to interfere with the patterns of exports of potatoes grown in Jersey to UK markets’.

(Note: Although the UK, the Channel Islands [including Jersey] and the Isle of Man were regarded as a single Member State [pre-Brexit, of course], the Court held that the provisions of the 2001 Act could act as a disincentive to exports of potatoes from Jersey to mainland UK – and hence could potentially restrict re-exports from the UK to other Member States. Thus the provisions of the 2001 Act fell within the ambit of EU law.)

14.9.2 ‘Indistinctly applicable’ rules

Where national rules are ‘indistinctly applicable’ MEQRs – that is, they do not distinguish between goods intended for the domestic market and goods intended for export – then there is simply no breach of Article 35. In Groenveld (1979), concerning Dutch rules banning sausage manufacturers in the Netherlands from using horse-meat (whether the sausages were intended for export or not), the ECJ stated that Article 35 did not apply.

The Court stated that Article 35 only:

These principles were confirmed in Jongeneel Kaas (1984). Dutch law regulated the quality and content of cheese produced in the Netherlands. The ECJ held that, as the national rules drew no distinction between the ultimate destination of the goods concerned, there was no breach of Article 35.

However, this does not mean that an indistinctly applicable measure can never breach Article 35. An example is Gysbrechts & Santurel Inter (Case C–205/07) [2008] ECR I–9947, the facts of which were given above. Here, the Court decided that, although a provision of national legislation may be indistinctly applicable in the sense that it applies to all goods, if ‘its actual effect is nonetheless greater on goods leaving the market of the exporting Member State than on the marketing of goods in the domestic market of that Member State’, then a prima facie breach of Article 35 has occurred. Such a measure was, however, capable of justification using Cassis de Dijon (1979) principles, such as consumer protection, as well as under Article 36, subject to satisfying the proportionality test.

SUMMARY

  • The ECJ defines goods very widely: ‘products which can be valued in money, and which are capable, as such, of forming the subject of commercial transactions’ (Commission v Italy). Narcotic drugs are not ‘goods’ (B.S. & C.A.).
  • Article 34 TFEU prohibits quantitative restrictions (QRs) on imports and all measures having equivalent effect (MEQRs). Article 34 is directly effective (Ianelli & Volpi v Meroni).
  • QRs are ‘measures which amount to a total or partial restraint of … imports’ (Geddo v Ente Nazionale Risi). QRs include import bans (R v Henn & Darby, French Turkeys, Conegate).
  • The ECJ has defined MEQRs on imports very widely: ‘All trading rules enacted by Member States which are capable of hindering, directly or indirectly, actually or potentially, intra-Community trade’ (Dassonville). As far as establishing an MEQR is concerned, it is irrelevant whether national rules apply only to imports (distinctly applicable measures) or apply equally to domestic goods and imports (indistinctly applicable measures).
  • Examples of MEQRs on imports: contents and ingredients restrictions (Cassis de Dijon, Sandoz, Beer Purity, Muller, Red Bull, Greenham & Abel); packaging requirements (Walter Rau v De Smedt, Mars); prohibitions on the use of goods (Toolex Alpha, Tinted film, Motorcycle trailers, Mickelsson & Roos); origin-marking requirements (Dassonville, Souvenir Jewellery); name restrictions (Smanor, Guimont); import inspections (Commission v France (Italian Wine)); authorisation requirements (Dynamic Medien).
  • Article 35 TFEU prohibits quantitative restrictions (QRs) on exports and all measures having equivalent effect (MEQRs).
  • ‘Quantitative restrictions’ on exports include export bans (R v Thompson & Others, Dusseldorp).
  • Distinctly applicable MEQRs on exports are prima facie prohibited (Bouhelier, Jersey Potatoes). Indistinctly applicable MEQRs on exports only breach Article 35 if they have as their ‘specific object or effect’ the restriction of exports (Groenveld, Gysbrechts).
  • Article 36 TFEU allows Member States to justify measures that restrict imports or exports on grounds of public morality, policy or security; protection of health and life of humans, animals or plants; protection of national treasures; protection of industrial and commercial property. It is a closed list. The burden of proof is on the national authorities to show (a) evidence that there is a risk and (b) that the restrictions imposed are ‘proportionate’ (ATRAL).
  • Public morality is something for Member States to decide in accordance with their own values (R v Henn & Darby).
  • Public policy was invoked in R v Thompson & Others and Ahokainen & Leppik.
  • Public security was successfully invoked in Campus Oil. It covers both internal and external security (Richardt & Les Accessoires Scientifiques).
  • Human health protection is the most important derogation under Article 36 (Toolex Alpha). Justification ‘must be based on a detailed assessment of the risk to public health, based on the most reliable scientific data available and the most recent results of international research’ (Greenham & Abel; B.S. & C.A.). Flexibility is permitted in situations of ‘scientific uncertainty’ (Sandoz).
  • Even if a measure is covered by Article 36, it will still be unlawful if it is ‘arbitrary’, which will be the case if there is already a lawful trade in similar items in the importing state (Conegate).
  • Member States must not use Article 36 as a ‘disguise’ for economic protectionism (French Turkeys).
  • In Cassis de Dijon, the Court introduced two key principles: (1) The ‘rule of reason’: Member States may maintain or impose trade barriers where ‘necessary’ to satisfy ‘mandatory requirements’; (2) the ‘rule of mutual recognition’: a presumption that goods lawfully sold in one Member State should be available in all others. This presumption is rebuttable, typically by referring to national characteristics, but the burden of doing so falls on the Member State which is resisting imports. The Cassis principles apply to both import and export restrictions (Dusseldorp, Gysbrechts).
  • The list of mandatory (‘imperative’) requirements is open ended. Examples include: consumer protection (Walter Rau v De Smedt), the success of which depends on the ‘presumed expectations’ of the ‘average … reasonably well-informed … reasonably observant and circumspect’ consumer (Mars, Clinique); environmental protection (Danish Bottles, Walloon Waste, Dusseldorp, Aher-Waggon, Radlberger & Spitz, Mickelsson & Roos); cultural protection (Cinéthèque); protection of fundamental rights (Schmidberger v Austria); road safety (Snellers Autos, Motorcycle Trailers); protection of children (Dynamic Medien); the fight against crime (Tinted Film).
  • In principle, the ‘rule of reason’ is limited to ‘indistinctly applicable’ measures (Gilli & Andres, Souvenir Jewellery). However, there is an apparent exception for environmental protection (Walloon Waste (import ban) and Dusseldorp (export ban)).
  • All derogations are subject to a test of ‘proportionality’. The test is failed if there is another method (e.g. clear labelling) capable of achieving the same objective which is less restrictive of trade (Walter Rau).
  • In Keck & Mithouard, the Court decided that ‘Contrary to what has previously been decided … certain selling arrangements’ were exempt from Article 34.
  • Examples include rules on pricing (Keck & Mithouard, Belgapom, Colruyt), Sunday trading laws (Punto Casa, Pelckmans Turnhout), rules on shops’ closing times (Tankstation t’Heukske), restrictions on which shops can sell goods (Processed Milk), rules on methods of sale (Burmanjer & Others – no itinerant sales; A-Punkt Schmuckhandels – no door-to-door sales; DocMorris – no cash or promotional gift incentives), advertising restrictions (Hünermund, Leclerc-Siplec).
  • Selling arrangements are exempt from Article 34, provided they ‘apply to all relevant traders’ and ‘affect in the same manner, in law and in fact, the marketing of domestic products and of those from other Member States’.
  • Total advertising bans are not selling arrangements because they have a greater impact (in fact) on the marketing of imported goods (Gourmet International Products, Douwe Egberts). The same is true of legislation banning Internet sales (Deutsche Apothekerverband, Ker-Optika). Such rules are MEQRs on imports, subject to Article 34.

A flow diagram depicts the means of establishing breach of Article thirty-four with ‘yes’ or ‘no’ conditions.

Long Description for Figure 14.1

The flow begins with the question, “Is the member state preventing free movements of goods?”. If ‘no,’ then “Article thirty-four or thirty-five will not apply,” and if ‘yes,’ then “Is the measure quantitative restrictions or measures equivalent to quantitative restrictions?”. If ‘no,’ then “Is the measure an indistinctly applicable measure equivalent to quantitative restrictions?”, if ‘yes, then “Is the measure a dual burden rule?” and if ‘no,’ then “Is the measure an equal burden rule?” and ‘yes,’ then “Articles thirty-four or thirty-five are not generally offended,” and if ‘no,’ then “Articles thirty-four or thirty-five apply, and the Cassie de Dijon rule of reasons subject also to second rule on mutual recognition. If “measure quantitative restrictions or measures equivalent to quantitative restrictions” is ‘yes,’ then “Does any Article thirty-six exemption apply?”. If ‘yes,’ then “The measure is lawful and there is no breach of Article thirty-four or thirty-five” and if ‘no,’ then “The measure is a breach of Article thirty-four or thirty-five”.

Figure 14.1 The means of establishing breach of Article 34

Further reading

Articles

  • Connor, T, ‘Accentuating the Positive: The “Selling Arrangement”, the First Decade and Beyond’ (2005) 54 ICLQ 127.
  • Davies, G, ‘Can Selling Arrangements be Harmonised?’ (2005) 30 EL Rev 370.
  • Hojnik, J, ‘Free Movement of Goods in a Labyrinth: Can Buy Irish Survive the Crises?’ (2012) 49 CML Rev 291.
  • Józon, M, ‘The Enlarged EU and Mandatory Requirements’ (2005) 11 ELJ 549.
  • Kaczorowska, A, ‘Gourmet Can Have His Keck and Eat It!’ (2004) 10 ELJ 479.
  • Krenn, C, ‘A Missing Piece in the Horizontal Effect “Jigsaw”: Horizontal Direct Effect and the Free Movement of Goods’ (2012) 49 CML Rev 177.
  • Notaro, N, ‘The New Generation Case Law on Trade and Environment’ (2000) 25 EL Rev 467.
  • Snell, J, ‘The Notion of Market Access: A Concept or a Slogan?’ (2010) 47 CML Rev 437.
  • Szydlo, M, ‘Export Restrictions within the Structure of Free Movement of Goods: Reconsideration of an Old Paradigm’ (2010) 47 CML Rev 753.

15 Article 28 TFEU and Customs tariffs and Article 110 TFEU and discriminatory internal taxation

DOI: 10.4324/9781003218562-15

15.1 The Common Customs Policy

In the previous chapter, the provisions of the TFEU dealing with ‘quantitative restrictions’ and ‘measures having an equivalent effect’ were examined. Articles 34 and 35 essentially deal with obstacles to trade caused by national legislation regulating the composition and/or packaging of goods. Articles 34 and 35 do not, however, apply to the situation when Customs officers in one Member State impose some form of financial charge on goods because they are crossing the border. Such charges are clearly capable of hindering trade, but they are tackled by different provisions of the TFEU. The key provision is Article 28, which creates the Customs Union. It has two parts:

  • A prohibition on Customs duties and charges having equivalent effect. This deals with the imposition of charges on goods moving around the EU from one Member State to another. This prohibition on Customs duties and equivalent charges is reiterated in Article 30.
  • A Common Customs Tariff. This deals with the movement of goods into the EU from elsewhere in the world, e.g. from China or the USA.

15.2 Article 30 and prohibition of Customs duties and charges having equivalent effect

15.2.1 Introduction

Article 30 prohibits Customs duties or charges having equivalent effect on all imports and exports between Member States. The ECJ has clarified that fiscal barriers to trade must be dealt with under Article 30 and not Article 34. In Compagnie Commerciale de l’Ouest and Others (Cases C–78 to 83/90) [1992] ECR I–1847, the Court stated:

A ‘Customs duty’ is any charge of a fiscal nature that is imposed, directly or indirectly, on goods which cross a border. A simple prohibition of ‘Customs duties’ per se would have allowed Customs authorities to continue to charge importers through less obvious means. Hence, Article 30 also prohibits charges having an effect equivalent to Customs duties (CEE). In Commission v Italy (Statistical Levy) (Case 24/68) [1969] ECR 193, a CEE was described as:

In Commission v Luxembourg and Belgium (Gingerbread) (Cases 2 and 3/62) [1962] ECR 813, the ECJ emphasised that it is the effect of a charge, as opposed to its name, which is significant. Otherwise, Member States could disguise charges on imports easily, and use them to destroy any competitive advantage which cheaper imports would otherwise have over domestic products. Such charges need not necessarily be levied at a frontier. The key is whether the charge has been imposed because the goods were imported (Steinike & Weinleg (Case 78/76) [1977] ECR 595). In Deutsches Milch-Kontor (Case C–272/95) [1997] ECR I–1905, the ECJ expressly held that Article 30 applied even though the charges in question were in respect of Customs inspections carried out within Germany on lorries heading for Italy.

Many ‘Customs duties’ are imposed to make imports relatively more expensive and hence protect the domestic market. But there is no requirement that the charge be levied for protectionist reasons (Commission v Italy (Statistical Levy) (1969)). There may even be no domestic market in need of protection; but this will not stop the charge from infringing Article 30. This was seen in Sociaal Fonds voor de Diamantarbeiders v SA Charles Brachfeld & Sons and Chougal Diamond Co (Cases 2 and 3/69) [1969] ECR 211. Here, Belgian Customs officials imposed a charge on uncut diamonds imported into Belgium. There are obviously no Belgian diamond mines (the diamonds originated in southern Africa), but the ECJ nonetheless held that Article 30 applied.

15.2.2 Derogations from Article 30

Charge for services for the benefit of the importer/exporter

Even if the charge is for a service provided for the benefit of the importer or exporter, it may still breach Article 30. A charge for services which benefits the Union in general, e.g. health inspections or quality control, would not necessarily be compatible with Article 30 either (Rewe-Zentralfinanz (Case 39/73) [1973] ECR 1039). To escape Article 30, a charge must be capable of being regarded as a payment for services of tangible benefit to the importer: see Commission v Italy (Statistical Levy) (1969). The charge must not exceed the value or cost of the service (Donner (Case 39/82) [1983] ECR 19), nor a sum proportionate to the service provided (Commission v Denmark (Case 158/82) [1983] ECR 3573). A charge based on the value of the goods is not permissible (Ford España v Spain (Case 170/88) [1989] ECR 2305).

Charges for services imposed under EU law

Where the services charged for are imposed under EU law, then a charge could be regarded as a payment for services and therefore it would be permissible for the Member State concerned to require payment for it (Commission v Italy (Statistical Levy) (1969)). Similarly, where a service is mandatory under international agreement, a charge could be levied to cover the cost of the service (Netherlands v P Bakker Hillegom (Case C–111/89) [1990] ECR I–1735). Where the state is entitled to recover its service costs through a charge, it is only entitled to recover the actual cost of the service, and no more (Denkavit Futtermittel v Germany (Case 233/81) [1982] ECR 2933). Where a service is only permitted this is insufficient (Commission v Belgium (Health Inspection Charges) (Case 314/82) [1984] ECR 1543). It is essential that the service is imposed under EU law (Germany v Deutsches Milch-Kontor (Case C–426/92) [1994] ECR I–2757).

No other exceptions permitted

There is no correlation with Article 36 or the Cassis de Dijon (1979) principles (see Chapter 14). In Commission v Italy (Export Tax on Art Treasures) (Case 7/68) [1968] ECR 617, the Italian government argued, unsuccessfully, that a charge levied on those exporting works of art from Italy was required to protect artistic heritage. In Diamantarbeiders (1969), above, the Court rejected any suggestion that the social utility of the Belgian Customs charge provided a defence (it had been argued that any money generated by the import charge on the imported diamonds would be used to provide financial aid for diamond mine workers in Africa). This case was followed in Kapniki Mikhailidis (Cases C–441 and 442/98) [2000] ECR I–7415, involving the imposition of charges on exports of tobacco products from Greece, apparently designed to raise money for workers in the tobacco industry. The Court held that this was no justification for breaching Article 30.

15.2.3 Repayment of illegal duties and charges

In several cases examined above it was established that a Customs duty or equivalent charge had been imposed by the customs authorities of a Member State in breach of Article 30. When that happens, then the Member State in question is, in principle, obliged to repay the person subject to the charge, usually the importer or exporter (San Giorgio (Case 199/82) [1983] ECR 3595; Dilexport (Case C–343/96) [1999] ECR I–579). Moreover, national rules of evidence which have the effect of making it virtually impossible or excessively difficult to secure repayment of charges levied in breach of Article 30 are incompatible with EU law. However, EU law does not require the repayment of the duty or charge in circumstances where this would unjustly enrich the person concerned (Just v Danish Ministry for Fiscal Affairs (Case 68/79) [1980] ECR 501; Kapniki Mikhailidis (2000)). This would occur in a situation, for example, where the burden of the charge has been transferred in whole or in part to other persons. Thus, if an importer had paid the charge, but then passed on the cost to the distributor of the goods in the importing state, then to order the repayment of the charge to the importer would over-compensate that person (Société Comatelo and others (Case C–192/95) [1997] ECR I–165).

15.3 The Common Customs Tariff

Article 28 also provides for the creation of a Common Customs Tariff (CCT), which co-ordinates the duties imposed on to all goods imported into the EU. A single tariff ‘wall’ is erected against imports, which no state is free to breach. The CCT came into operation in July 1968. It lays down common rules on nomenclature, valuation and origin. CCT duties are fixed by the Council, following proposals made by the European Commission (Article 31).

15.4 Article 110 and the prohibition of discriminatory internal taxation

15.4.1 The scope of Article 110

Article 110 allows Member States the freedom to establish their own taxation system for any given product, provided that there is no discrimination against imports, or indirect protection of domestic products. The purpose of Article 110 is to remove discrimination against imports, not to accord them tax privileges (Kupferberg (Case 253/83) [1985] ECR 157). Therefore, internal taxation may legitimately be applied to a particular product even if there is no domestic production of that product, and the tax is effectively levied against imports only. However, this is permitted because the Member State is not discriminating against imports. It would only be discriminatory if it was imposing a different rate of tax on similar or competing domestic products, but if there are no similar domestic products, no breach of Article 110(1) can occur, and if there are no competing domestic products, no breach of Article 110(2) can occur either. This was seen in De Danske Bilimportører (Case C–383/01) [2003] ECR I–6065. A Danish company had bought and imported a new German-made Audi car. Under Danish law, all new cars must be registered and a ‘registration duty’ paid. The purchase price of the car was approximately €27,000; the duty came to just over €40,000, bringing the total price to some €67,000. The company could not believe that this was correct and brought a legal challenge. The ECJ decided that no breach of Article 110 had occurred, simply because there were no similar (or even competing) domestic products – Denmark does not manufacture cars. Finally, because Article 110 prohibits discrimination against imports, it does not preclude the imposition of higher rates of tax on domestic products than on imports: that is, discrimination in favour of imports (known as reverse discrimination) is permitted (Grandes Distilleries Peureux (Case 86/78) [1979] ECR 897).

It is very important to remember that internal taxation is only prohibited by Article 110 if it is either discriminatory (against imports) or has a protective effect (in favour of domestic goods). To a very large extent, therefore, the 27 Member States of the European Union are allowed complete discretion whether or not to impose taxes – or ‘excise duties’ – on a range of goods and, if they do impose such taxes, at what rate. This can lead to significant differences in prices from one Member State to another.

‘Products of other Member States’

Despite the clear implication of this phrase, the ECJ has held that the prohibition of discriminatory taxation must apply to goods manufactured or produced outside the EU, but which are in free circulation inside the EU (Co-operativa Co-frutta (Case 193/85) [1987] ECR 2085). Using even more expansive interpretation, the ECJ has held that Article 110 applies to exports, in order to guarantee the neutrality of national systems of taxation (Larsen (Case 142/77) [1978] ECR 1543). However, there are some limits. In Air Liquide (Cases C–393/04 and C–41/05) [2006] ECR I–5293, involving a Belgian tax on the use of industrial, commercial, financial or agricultural motors, the Court held that Article 110 was inapplicable, because there was no tax on ‘products’ as such. The Court stated that the ‘tax on motive force … is not imposed specifically on exported or imported products. It applies to economic activities carried out by industrial, commercial, financial or agricultural undertakings and not to products as such’.

15.4.2 Distinguishing Customs duties and taxes

Charges imposed only on imports are clearly contrary to the provisions on Customs duties and are dealt with under Article 30. The position is different where a charge is applied to all goods: domestic products as well as imports. A charge levied indiscriminately on all products of a particular description, in all respects, regardless of the country of origin, is a tax, not a Customs duty, and should be dealt with under Article 110. The picture can be complicated because many Customs duties are disguised as taxes (e.g. Commission v Luxembourg and Belgium (Gingerbread) (1962); Orgacom (Case C–254/13) [2015] 1 CMLR 27).

It is important to distinguish genuine taxes from disguised Customs duties. A ‘tax’ was defined in Commission v France (Reprographic Machines) (Case 90/79) [1981] ECR 283 as one which related to a ‘general system of internal duties applied systematically to categories of products in accordance with objective criteria irrespective of the origin of the products’. It follows that a charge which is apparently levied on all goods but which is imposed in a different manner, and/or which is subject to different criteria, for domestic and imported products, would fall to be dealt with under Article 30 (Marimex (Case 29/72) [1972] ECR 1309; Orgacom (2015)).

Moreover, ‘taxes’ which actually benefit domestic products are, in reality, Customs duties. In Capolongo (Case 77/72) [1973] ECR 611, a ‘tax’ was imposed under Italian law on all cellulose products, such as egg boxes. The proceeds were used to finance the production of paper and cardboard in Italy. The ECJ held that such a ‘tax’ may fall foul of Article 30 if the proceeds of the charge were used to benefit a domestic product. The Court went on to hold that this applied even where the product on which the charge was levied and the domestic product which benefited from it were different.

However, the very wide principle established in Capolongo (1973) has subsequently been restricted. In Fratelli Cucchi (Case 77/76) [1977] ECR 987, involving an Italian ‘tax’ on sugar, the ECJ held that a ‘tax’ would only constitute a CEE in breach of Article 30 if certain criteria were satisfied:

  • the tax would need to have ‘the sole purpose of financing activities for the specific advantage of the taxed domestic product’;
  • the taxed product and the domestic product benefiting from the charge must be the same;
  • taxes imposed on the domestic product would need to be made good in full.

Thus, the charge that was the subject of the Capolongo (1973) litigation would not now be found to be in breach of Article 30, because the products involved were different. However, such a charge could very well be regarded as a discriminatory tax instead, in breach of Article 110 (Co-operativa Co-frutta (1987), concerning an Italian ‘tax’ on bananas). The Fratelli Cucchi (1977) criteria have been confirmed in a succession of cases subsequently. In UCAL (Case C–347/95) [1997] ECR I–4911 and Fricarnes (Case C–28/96) [1997] ECR I–4939, involving Portuguese ‘taxes’ on dairy products and meat, respectively, the position was summarised as follows:

15.4.3 Discrimination against imports: Article 110(1)

The ECJ has held that Article 110(1) must be construed broadly, to include all taxation actually and specifically imposed on a particular product. Article 110(1) prohibits the imposition of taxation on domestic products which discriminates against similar imported products. Discrimination may either be direct or indirect and still breach Article 110(1) – the difference is that indirect discrimination may be justified.

Direct discrimination

Discrimination may be ‘direct’, that is, the discrimination is visible on its face. Alfons Lütticke (Case 57/65) [1966] ECR 293 involved German tax legislation that discriminated against the same type of goods – dried milk – purely according to its country of origin. This clearly breaches Article 110(1). Moreover, in Haahr Petroleum (Case C–90/94) [1997] ECR I–4085, the ECJ rejected any suggestion that directly discriminatory taxation could be justified.

Indirect discrimination

Where a taxation system is prima facie non-discriminatory, but in practice discriminatory against imported products, this is known as indirect discrimination. It will nevertheless amount to a breach of Article 110(1). This is perfectly demonstrated in Humblot (Case 112/84) [1987] ECR 1367, a case involving French road tax.

Note: the French franc was abolished when the single European currency, the euro, came into operation in 2002.

Indirect discrimination, unlike direct discrimination, may be justified, provided that the difference in treatment of domestic goods and imported goods is designed to achieve some acceptable outcome. The difference must be based upon objective criteria. In Commission v France (Tax on Wines) (Case 196/85) [1987] ECR 1597, sweet wines were taxed at a lower rate than ordinary wines. This was done in order to provide some economic assistance to rural areas in France that were dependent on sweet wine production. The ECJ held that the tax was prima facie in breach of Article 110(1) but, as the discrimination was indirect, it was capable of justification.

‘Similar products’

It is not necessary that the imported products and the domestic products which are subjected to different rates of taxation are identical. They need only be ‘similar’, and in Commission v France (Taxation of Spirits) (Case 168/78) [1980] ECR 347, the ECJ held this word should be construed broadly.

The Court held:

Most cases have arisen in the context of alcoholic drinks, primarily because they are taxed so heavily by most Member States’ governments for social and/or revenue-raising reasons. There are also many fine distinctions that can be drawn between different drinks, depending on their raw ingredients, strength, method of manufacture, etc.:

  • Commission v Denmark (Tax on Wines) (Case 106/84) [1986] ECR 833 – wine and fruit wine were ‘similar’;
  • John Walker & Sons (Case 243/84) [1986] ECR 833 – whisky and fruit liqueur wines were not ‘similar’, because whisky contained twice as much alcohol (40% to 20%).

FG Roders BV (Cases C–367 to 377/93) [1995] ECR I–2229 involved a very thorough examination of the similarity of a number of drinks, including French red wine and champagne, Italian vermouth, Portuguese madeira and Spanish sherry. BENELUX legislation (meaning legislation common to Belgium, Luxembourg and the Netherlands) taxed wines differently according to whether they were still or sparkling, and according to whether they had been made from grapes or from fruit. There were therefore four categories into which wine could fall for tax purposes:

  • still grape wine;
  • still fruit wine;
  • sparkling grape wine; and
  • sparkling fruit wine.

Still fruit wine was exempt from tax altogether (provided that it met certain requirements with regard to labelling and packaging) while sparkling fruit wine was taxed at a lower rate than sparkling grape wine. Various importers into the Netherlands of French red wine and champagne, Italian vermouth, Portuguese madeira and Spanish sherry (all of which are made from grapes) argued that their products were ‘similar’ to domestically produced fruit wine (whether still or sparkling, as the case may be) and should therefore be taxed at the lower rate. The Court made a number of findings, looking at each of the imported drinks in turn.

  1. Red table wine is ‘similar’ to still fruit wine. It was in this specific context that the Court made its finding:

    the two categories of wine possess similar organoleptic properties – in particular taste and alcohol strength – and meet the same needs of consumers, inasmuch as they can be consumed for the same purposes, namely both to quench thirst and to refresh, and to accompany meals.

  2. Red quality wine may be ‘similar’ to still fruit wine, but this was a question for the national court, by examining the ‘objective characteristics of both categories of beverage, such as their origin, their method of manufacture and their organoleptic properties, in particular taste and alcohol content’, and, second, by considering ‘whether or not both categories of beverage are capable of meeting the same needs from the point of view of consumers’.

  3. Vermouth may also be ‘similar’ to still fruit wine, although again this was a question for the national court. However, the Court did point out two potentially significant differences: ‘not only is ethyl alcohol added to grape wine but also a small quantity of mixed herbs which give vermouth its special flavour’.

  4. Champagne may also be ‘similar’ to sparkling fruit wine, although that would be for the national court to decide. Again, the Court pointed out a number of differences between the two drinks:

    champagne is made sparkling by a natural method – by a second alcoholic fermentation in the bottle [but] sparkling fruit wines require the addition of carbon dioxide, a fermentation process which is not natural. Secondly, the organoleptic properties of champagne are not comparable to those of sparkling fruit wines. Thirdly, the two categories of beverage do not meet the same needs of consumers, particularly since the consumption of champagne is usually associated with special occasions.

  5. Finally, sherry and madeira are not ‘similar’ to still fruit wines. There were two reasons for this; first, that the imported drinks were ‘usually consumed as aperitifs or as dessert wines’ and, second, because they had a higher alcohol content (17%–18%) than fruit wines (15%).

15.4.4 Indirect protection of domestic products: Article 110(2)

It is unnecessary for the products in question to be ‘similar’ for the purposes of Article 110(2). Instead, it applies to ‘all forms of indirect tax protection in the case of products which, without being similar within the meaning of [Article 110(1)], are nevertheless in competition, even partial, indirect or potential, with each other’ (Co-operativa Co-frutta (1987)). It is thus much wider than Article 110(1). In several of the cases considered above under Article 110(1), the ECJ found that the products in question were not ‘similar’. However, the Court then went on to discuss whether the dissimilar products may, instead, be ‘in competition’ with each other so that differential levels of taxation may have a protective effect for domestically produced goods. Hence:

  • Commission v France (Taxation of Spirits) (1980) – brandy and cognac (domestically produced in France) and whisky and gin (imported) were not ‘similar’ to but were in competition.
  • Commission v UK (Tax on Beer and Wine) (Case 170/78) [1983] ECR 2265 – beer (domestically produced in the UK) and wine (imported) were not ‘similar’ but beer was in competition with at least some wines.
  • Commission v Italy (Tax on Fruit) (Case 184/85) [1987] ECR 2013 and Co-operativa Co-frutta (1987) – bananas (imported) were not ‘similar’ to other fruits (domestically grown in Italy) but were in competition with them.
  • Commission v Sweden (Tax on Beer and Wine) (Case C–167/05) [2008] ECR I–2127 – strong beer (produced in Sweden) was in competition with at least some wines (imported).

It should be noted that, just because dissimilar products are in competition with each other, it does not mean that taxing one at a higher rate will have a protective effect on the other. It may be that the imported product is of such a significantly higher quality such that taxing it will have little or no impact on sales. Alternatively, as was the case in Commission v Belgium (Tax on Wine and Beer) (Case 356/85) [1987] ECR 3299, if the price differential between the imported product and the dissimilar domestically produced product is already quite substantial, differences in the rates of tax will not serve to protect the domestically produced product.

Finally, in Commission v Sweden (2008), the Court found that the difference in price between Swedish beer and imported wine was virtually the same after taxation as it was beforehand, and hence the fact that Sweden taxed wine more heavily than beer was not ‘likely to influence consumer behaviour’. In other words, the tax system did not protect the domestic product (beer) against competition from the imported product (wine).

SUMMARY

  • Article 30 TFEU prohibits ‘Customs duties on imports and exports and Charges having equivalent effect.’
  • Equivalent charges are defined very widely. All charges imposed on goods when they cross a national frontier, whether by the importing or exporting state, are prohibited. The size, and the name, of the charge are immaterial (Commission v Italy (Statistical Levy)).
  • It is irrelevant why a Customs duty or equivalent charge has been imposed. It may have been for socially valid reasons (e.g. raising cash for disadvantaged workers) but this is immaterial (Sociaal Fonds, Kapniki Mikhailidis).
  • Article 110 TFEU allows Members States to establish their own ‘internal taxation’ system, in the form of excise duties on goods, but prohibits discrimination against imported products. Discriminatory excise duties are prohibited whether direct (Alfons Lütticke) or indirect (Humblot). Direct discrimination cannot be justified (Haahr Petroleum), but indirect discrimination is justifiable (Commission v France (Tax on Wines)). ‘Excise duties’ are most frequently imposed on tobacco products, alcoholic drinks and fuel.
  • ‘Internal taxation’ means a ‘general system of internal duties applied systematically to categories of products in accordance with objective criteria irrespective of the origin of the products’ (Commission v France (Reprographic Machines)).
  • A so-called tax which is apparently levied on all goods but which is imposed in a different manner, and/or which is subject to different criteria, for domestic and imported products, is therefore not a ‘tax’ and falls under Article 30 instead (Marimex). Similarly, a so-called tax which actually benefits domestic products is not a ‘tax’ either and falls under Article 30 instead (Capolongo). However, the products in question must be the same (Fratelli Cucchi).
  • Imported goods must not be taxed ‘in excess’ of the taxation imposed on ‘similar’ domestic products (Article 110(1)). ‘Similar’ products are ‘products which have similar characteristics and meet the same needs from the point of view of consumers’ (Commission v France (Tax on Spirits)). Hence products do not have to be identical for Article 110(1) to apply.
  • Where different products are in ‘competition’ with each other, Member States may impose different excise duties but the level of taxation imposed on imported products must not afford ‘indirect protection’ to domestic products (Article 110(2)). Products need not be ‘similar’ for Article 110(2) to apply but must be in ‘competition, even partial, indirect or potential, with each other’ (Co-operativa Co-frutta). Beer and wine are not ‘similar’ products, but they are at least in ‘competition’ with each other (Commission v UK (Tax on Beer and Wine)).
  • Article 30 applies when imports are subject to a charge even if there are no competing domestic products (Sociaal Fonds). However, for Article 110 to apply, there must be ‘similar’ or at least competing domestic goods (De Danske Bilimportører).

Further reading

Articles

  • Easson, A, ‘The Spirits, Wine and Beer Judgments: A Legal Mickey Finn?’ (1980) 5 EL Rev 318.
  • Easson, A, ‘Fiscal Discrimination: New Perspectives on Article 95’ (1981) 18 CML Rev 521.
  • Lonbay, J, ‘A Review of Recent Tax Cases’ (1989) 14 EL Rev 48.
  • Plender, R, ‘Charges Having an Equivalent Effect to Customs Duties: A Review of the Cases’ (1978) 3 EL Rev 101.
  • Snell, J, ‘Non-discriminatory Tax Obstacles in Community Law’ (2007) 56 ICLQ 339.

16 EU competition law

DOI: 10.4324/9781003218562-16

16.1 The basis of EU competition law

16.1.1 The purpose of competition law

Article 2 EC Treaty originally identified the main task of EU law as being:

As well as this Article 3(g) also states that the EU must ensure:

As a result, the framers of the EU Treaty included very specific rules regarding competition. To have rules on competitive, or rather anti-competitive, behaviour seems at first sight to be inconsistent with the whole idea of economic freedom that is demonstrated in the ‘Four Freedoms’ and the drive towards a Single Market unfettered by any national barriers to trade. Rules on competition would seem to be a restriction on business efficiency and a restriction on successful businesses and therefore on market forces. This is in fact not the case. The Treaty makers were not trying to restrict businesses from free competition but rather to prevent large powerful businesses from using unfair means to harm small- and medium-sized businesses.

The drafters of the Treaty realised that within a newly created Single Market companies would enjoy the benefit of a free and expanded market and might see this as a chance to get round former national policies on anti-competitive activities and operate on a much larger scale to maximise their potential. In other words, the framers of the Treaty recognised the point that unfettered market forces can lead directly on to monopolies. (A monopoly is where a single company or organisation enjoys almost unrestricted trade in a particular product or service.) Monopolies, almost by definition, defeat competition and ultimately, therefore, restrict consumer choice.

As a result, the framers of the Treaty inserted rules in the Treaty aimed at dealing with such practices. In doing so they took the same view as expressed in the Sherman Act of 1890 of the United States of America, that there should be prohibition of ‘every contract, combination or conspiracy in restraint of trade’ and also ‘the monopolization of trade and commerce’. In other words, the rules should prevent the distortion of free competition resulting from either collusion between different businesses or the predominant power of a single business.

Therefore, the framers of the Treaty then drafted in a series of rules with three principal objectives:

  • to avoid the possibility of restrictive practices and agreements;
  • to prevent large businesses from abusing their economic dominance in the market;
  • to ensure that, within certain limits, the public sector observes the same rules.

Set against the problems foreseen when framing the Treaty, these seem to be quite reasonable controls. Nevertheless, EU competition law has often been criticised in a number of different ways.

First, it has been said that the law is too general in its application. This is because all agreements and practices are treated alike, even if they would actually benefit the consumer.

Second, competition law has been criticised on the basis that it in fact only sacrifices equity and efficiency to political goals.

[The existence of competition law] is ‘a perversity’ and ‘the loud admission of defeat’ and ‘an admission that the market alone cannot effect competition’.

Another danger, of course, is always that ‘member states will continue to promote the national interest over that of the Community, cling rigorously to currency, sovereignty, and intensify economic divergences within the Community’.

I Ward, A Critical Introduction to European Law (Butterworths, 2003), p. 126

Nevertheless, despite the apparent shortcomings, it is generally accepted that the EU institutions have succeeded in framing a cohesive and consistent set of objectives to be pursued in implementing competition policy. These are all identified in the Commission’s Ninth Report on Competition Policy:

  • to create an open and unified market which is not partitioned by restrictive and anti-competitive agreements between firms;
  • to realise an appropriate amount of effective competition in markets, avoiding over-concentration or any abuses exercised by dominant companies;
  • to achieve fairness in the marketplace, which involves giving support to small and medium-sized firms, measures for the protection of consumers and the penalising of unlawful state subsidies;
  • to maintain the competitive position of the Community against its principal rivals in global economy, being mainly the USA and Japan.

16.1.2 The character of competition law

The rules on competition law cover all items capable of forming the subject matter of commercial transactions. So it includes not only goods and services, but also intellectual property rights. The provisions are actually framed in quite broad terms and as a result have been the subject of much interpretation in the ECJ.

The competition rules can be used against firms regardless of the existence of a registered office in the EU, provided that its actions are capable of affecting trade within the Single Market. This was clearly identified in The Woodpulp case, Ahlstrom Osakeyhtio and Others v Commission (Cases 89, 104, 114, 116, 117 and 125–129/85) [1993] ECR I–1307. The rules are also pragmatic in order to preserve free trade. In this way they are subject to exceptions to preserve market efficiency.

EU law only requires that penalties should apply where the actions of ‘undertakings’ affect trade between Member States. Nevertheless, this means that they can apply even if the effect is actually quite small, and even if the undertakings involved are in the same Member State.

It is also possible, therefore, for EU competition law and national law to exist alongside each other.

Articles 101 and 102 in fact complement each other, in that they pursue common objectives but by focusing on different types of activity in different situations.

It is evident, then, that EU competition law has three main objectives which may at times, nevertheless, appear to be incompatible:

  • efficiency;
  • protection of both consumers and small businesses;
  • the creation of the Single Market.

It is also true to say that competition law is among the most highly developed of all EU law. In fact, as Josephine Steiner points out:

So pervasive is the influence of EU competition law and so severe are the sanctions for its breach, that business, whatever its size, whether or not their operations are currently confined to their domestic market, cannot afford to ignore Community law.

J Steiner, Textbook on EC Law (2nd edn, Blackstones, 1991), p. 106

16.2 Article 101 and provisions on restrictive practices

16.2.1 The scope of Article 101

Article 101 is concerned with restrictive trade practices. It sets out distinctly the nature of the prohibition:

Article 101(1) also goes on to identify specific examples of anti-competitive acts that would fall under the prohibition as anti-competitive practices. These are those which:

  • Directly or indirectly fix purchase or selling prices or any other trading conditions. This covers any arrangement that directly or indirectly could hinder intra-Union trade. In this way it can include arrangements whereby undertakings agree on trading conditions applicable to their business dealings such as discounts or credit arrangements. It has, for instance, included a retail price maintenance agreement between Belgian and Dutch booksellers (VBVB and VVVB v Commission (Cases 43 and 63/82) [1984] ECR 19).
  • Limit or control production, markets, technical development, or investment. This involves agreements where undertakings restrict their own growth in order to raise prices artificially and prevent those outside the agreement from entering the trade, as in the Quinine Cartel case [1970] ECR 661 (see below).
  • Share markets or sources of supply. This refers to the situation where competitors agree to apportion markets on either a geographical or a product basis. This was in effect what was being done in Consten and Grundig [1966] ECR 429 (see below). It might be done in an oligopical market where competitors appoint each other as exclusive dealers of the other’s product or in a particular region of the EU.
  • Apply dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage. The object here clearly is to place the competing party at a disadvantage. It might obviously then involve providing advantageous conditions to one purchaser of a product over another.
  • Make the conclusion of contracts subject to acceptance by the other parties of supplementary obligations which, by their nature or according to commercial usage, have no connection with the subject of such contracts. This refers to situations where in order for a party to enter a contract it is bound to fulfil other obligations which in fact have no real bearing on the contract. A classic example would be requiring a party buying one product or service to buy at the same time a completely unrelated product or service.

Article 101(2) makes all such agreements void. It is then possible for the Commission to grant exemptions in certain cases.

The prohibition under Article 101 has three key elements all of which need to be proved and therefore all of which need to be defined and understood:

  • the types of agreements that are prohibited;
  • the effect on inter-state trade;
  • the meaning of ‘object or effect of preventing, distorting or restricting competition’.

Before looking at these it is important to understand the meaning of the word ‘undertaking’.

16.2.2 The concept of ‘undertaking’

Both Article 101 and Article 102 are concerned with the anti-competitive activities of ‘undertaking’. The term ‘undertaking’, while used widely, is not defined anywhere in the Treaty and so again we have to look to the case law for interpretation of the term by the ECJ.

In fact, the ECJ has given a broad definition to the term: ‘a single organisation of personal, tangible and intangible elements, attached to an autonomous legal entity and pursuing a long term economic aim’ (Mannesmann v High Authority (Case 19/61) [1962] ECR 357). The term has also been defined in numerous other cases and in Commission Reports on Competition Law.

As a result, it can be seen that the term covers almost every type of entity regardless of its legal status, from an individual to a multi-national corporation, provided that it has legal capacity and is engaged in an economic activity. Therefore there is a wide spread of disparate economic activities to which the term has been applied over time:

  • an opera singer (Re Unitel (Commission Decision 78/516) [1978] 3 CMLR 306);
  • a sports federation (Re World Cup 1990 Package Tours (Commission Decision 92/51));
  • a state-owned corporation (Italian State v Sacchi (Case 155/73) [1974] ECR 409);
  • a public agency (Höfner v Macrotron (Case C–41/90) [1991] ECR I–1979);
  • a trade association (FRUBO v Commission (Case 71/74) [1975] ECR 563);
  • banking (Zuchner v Bayerische Vereinsbank AG (Case 172/80) [1981] ECR 2021).

It has been held that the definition does not depend on the activity involving a profit motive. Nevertheless, the General Court has recently refined the definition and stated that there must be some form of economic activity, however marginal, for the entity to be regarded as an ‘undertaking’ for the purposes of competition law.

Throughout its jurisprudence, the ECJ reaffirms that where a not-for-profit seeking activity operating for a social purpose on the principle of solidarity is not an undertaking (Joined Cases C–264/01, C–306/01, C–354/01 and C–355/01 Aok Bundesverband and Others).

The issue of whether a sporting body can be classed as an undertaking has also been considered. The Bosman ruling, of course, suggested that this would only be the case where an economic activity was involved. It could clearly impact upon the law governing free movement of workers for instance where rules of sporting associations limited the numbers of non-national players. However, rules that merely deal with sporting conduct are not within the scope of EU law, and this point has been considered in relation to the International Olympic Committee.

It is clear in any case that the definition must be taken on a case-by-case basis.

16.2.3 The character of prohibited agreements

There are in fact three distinct types of agreement or restrictive practice which fall within the prohibition in Article 101:

  • agreements between undertakings;
  • decisions by associations of undertakings;
  • concerted practices.

Agreements between undertakings

An agreement between undertakings must always carry with it some form of collusion. This is inevitably to distinguish it from unilateral acts.

However, the agreement must also involve autonomous behaviour by the undertakings. For instance, Article 101 would not be infringed and it could not be classed as an agreement where national law imposed the agreement on the undertakings.

According to Tepea v Commission (Case 28/77) [1978] ECR 1391, it is clear that even informal, oral arrangements can be classed as an agreement for the purposes of Article 101. Indeed, while contractual arrangements would obviously be agreements there is no real requirement that the agreement should be binding for Article 101 to apply. In fact, the leading case has shown that the so-called gentlemen’s agreements could still fall foul of Article 101 and in any case the third type, concerted practices, would cover most collusive behaviour.

Decisions by associations of undertakings

Many industries collectively act within trading associations. Usually such associations co-ordinate behaviour and regulate standards in a trade and so incorporate a set of rules by which the members agree to be bound on membership. This is obviously usually beneficial. Nevertheless, when the rules of the association laid down are clearly aimed at harming free competition then they may amount to a breach of Article 101.

The obvious types of decision that could amount to a breach of Article 101 include ones fixing prices, or requiring specific discounts, or requiring collective boycotts of other undertakings, or the inclusion of any kind of restrictive contract clauses.

It has also been held, however, in NV IAZ International Belgium v Commission (1983) that even non-binding arrangements might count as a decision leading to a breach of Article 81 (now Article 101 TFEU). Neither does the body concerned have to be involved in commercial activity.

However, there will be no breach of Article 101 where trade associations are genuinely independent of their parent bodies.

Concerted practices

The term ‘concerted practices’ has been defined both by academics and the ECJ. Academics have defined it as ‘co-ordinated action between undertakings which, without amounting to an agreement, consciously substitutes co-operation for competition’. So the key characteristic of a concerted practice, by contrast with a straightforward agreement, is that it is more disguised or hidden and is generally informal.

In the leading case the ECJ has defined ‘concerted practice’ as:

So the true test of a concerted practice is when the parallel behaviour can be shown to be co-operative such that the undertakings involved appear to be acting with a common design or purpose: Co-operative Vereniging ‘Suiker Unie’ v Commission (the Sugar Cartel case) (Cases 40–48, 50, 54–56, 111, 113 and 114/73) [1975] ECR 1663. Even a single meeting can amount to a concerted practice (T-Mobile Netherlands BV (Case C–8/08) [2009] ECR I–4529).

Horizontal agreements and vertical agreements

Agreements that may offend Article 101 may of course be of two different types: horizontal or vertical.

A horizontal agreement is one between undertakings at the same level, so would usually be one between competing manufacturers or distributors. This could, for example, be an agreement to divide up markets; or it could be a price-fixing arrangement as in the Dyestuffs case (ICI v Commission) (1972).

A vertical agreement, on the other hand, is one reached between undertakings at different levels in the process, for example between the manufacturer and the distributor, or between wholesalers and retailers. These would usually, of course, benefit the consumer because they are likely to streamline the process of trade. However, they may offend Article 101 if they involve exclusive distribution arrangements, or exclusive licensing agreements.

The principles were established in the leading case, commonly referred to as Consten and Grundig.

16.2.4 The effect on trade between Member States

In order to establish that there has been a breach of Article 101 the Article also requires that the agreement may affect trade between Member States. The important word here is ‘may’. In other words, there is no absolute requirement that the agreement has in fact affected trade; merely that it has the potential to do so.

The capacity of the agreement to affect trade is ascertained by reference to the free movement of goods and attainment of a Single Market, the same basic test as that identified in the Dassonville formula (see section 14.4).

So an agreement is capable of affecting trade between Member States if it is capable of constituting a threat ‘direct or indirect, actual or potential, on the pattern of trade’.

In this way there is no need to prove any actual harm, as long as the agreement is likely to prevent, restrict or distort competition to a sufficient degree.

16.2.5 The object or effect of preventing, restricting or distorting competition

Article 101 requires also that the agreement must have as its ‘object or effect the prevention, restriction or distortion of competition within the common market’.

‘Object’ and ‘effect’ in the context of the Article are clearly meant to be alternative tests. On this basis the test has as much to do with the practical outcomes of a business arrangement as it has to do with the intentions of the undertakings that are party to the arrangement.

The key issue then is whether competition has been stifled or affected, rather than whether there has been any actual movement in trade, whether up or down.

However, the ECJ nevertheless tries to apply the rules in a way that will not stifle business enterprise and initiative. The concern is not to prevent businesses from operating efficiently and effectively but to prevent a distortion of real competition.

In the above case the ECJ identified the factors that will need to be taken into account to determine whether or not an agreement is capable of distorting competition:

  • The nature and quantity of the product in question – so that the greater the market share of the product concerned, the more likely it is that an agreement may inhibit competition.
  • The position and size of the undertakings involved – so that the greater the share of the market that they enjoy, the more there is a possibility of distorting competition.
  • The relationship of the agreement to other agreements – so that the more isolated the agreement, the less likely it is to limit competition, but where it forms part of a network of agreements it is more likely to affect competition.
  • The extent of the agreement – so that any agreement that extends beyond what is necessary to achieve the desired beneficial risk is likely to limit competition.
  • The link with agreements on parallel imports or exports – so that where the agreement also includes bans on parallel imports or exports it will most usually be seen as stifling competition and lead to a breach of Article 101.

Market definition is clearly an important question and is one that the Commission has addressed in the Notice on Definition of the Relevant Market (1997 OJ C372/5). It has also been considered in the case law.

Traditionally, in any case, the ECJ would apply the de minimis rule. The Court would not show interest in agreements which would have a disproportionately minor effect on competition and would only strike down agreements that might affect competition to a noticeable extent.

In fact, the de minimis principle has been introduced by the Commission in the form of Notices on Agreements of Minor Importance. The first of these was in 1986. The most recent is the Commission Notice on Agreements of Minor Importance 2001 (OJ 2001 C368/13).

Under this Notice an agreement will not be in breach of Article 101:

  • where the undertakings that are parties to the agreement are actual or potential competitors (in other words, in a horizontal agreement), the aggregate market share of the undertakings concerned does not exceed 10% of the market; or
  • where the undertakings that are parties to the agreement are not competitors (in other words, in a vertical agreement) the aggregate market share of the undertakings concerned does not exceed 15% of the market.

However, there are qualifications to these basic rules:

  • in horizontal agreements the agreement must not contain any restrictions on sale price, limitation of output, or allocation of markets or customers, otherwise the notice will not apply;
  • in vertical agreements the agreement must not contain restrictions on the minimum resale price of goods or on the territory that the goods will be sold in or the consumers that the goods will be sold to, or the notice will not apply.

16.2.6 Exemptions

Not all agreements will be automatically found to be in breach of Article 101. There are in fact different ways of avoiding being caught by Article 101 and until recently these included the granting of exemptions by the Commission. Since 1 May 2004, under the operation of Regulation 1/2003 a different process is in operation involving the Member States.

Sensibly, the Treaty also provided the means by which it could be identified that certain agreements could be exempted from the operation of Article 101. This includes both individual exemptions (which follow individual applications) and block exemptions that can be applied to specific categories of agreement.

Article 101 paragraph 3 specifically creates the criteria for exempting agreements. There are four conditions that must be met in order for exemption to be granted. Two of these are phrased in positive terms and two are phrased in negative terms.

  • First, it must be possible to show that the agreement, decision or practice contributes in some way to improving the production or distribution of goods or alternatively to promoting technical or economic progress.
  • Second, a fair share of the resulting benefit must pass to the consumer. This need not merely be the end consumer.
  • The agreement or practice must not impose any unnecessary restrictions that would go beyond the necessary positive aims. An example of a practice that would infringe this negative requirement and therefore mean exemption would be denied is an absolute territorial protection, such as the ban on parallel imports in Consten and Grundig v Commission (1966).
  • There must not be any possibility of the restrictions eliminating competition in respect of a substantial part of the product in question. Obviously, in respect of this requirement the market share of the parties concerned and the level of competition within the specific market will be crucial factors. In Re Vacuum Interrupters (1977) (above) the market stretched well beyond the EU because the undertakings faced significant competition from both the USA and Japan. However, where the competition is much more limited the market might be restricted to a single Member State.

Individual exemptions

Prior to 2004 individual exemptions were granted by the Commission under the procedure in Regulation 17/62. Application was by way of a ‘notification’ of the agreement or practice to the Commission. Individual exemption was then granted in the form of Decisions. These might be for only limited periods and they might also be conditional or depend on the fulfillment of certain criteria or certain obligations identified in the Decision.

Now, under Regulation 1/2003 it is no longer necessary to notify the Commission in order to obtain individual exemption. Article 1(1) of the Regulation instead places the burden on undertakings themselves to identify whether any of their agreements offend Article 101(1) but are in fact exempt under Article 101(3). The national authorities are then responsible for applying the criteria for determining whether a particular agreement is in fact exempt.

Unlike the former system where the Commission had to be notified for exemption to be granted there is no requirement to inform the national authorities in the same way for the exemption to take effect. National authorities will determine whether or not there has been an infringement of Article 101 by the agreement in question.

Block exemptions

Block exemptions were in any case a means of reducing the huge bureaucratic burden on the Commission of applications for individual exemptions. The process was introduced then for the purpose of streamlining competition law.

Inevitably, also, the process of making application for individual exemption left businesses in a state of uncertainty until a Decision was issued. On this basis the use of block exemptions allowed businesses to assess for themselves whether or not a particular type of agreement was exempt. If it was apparent that a particular type of agreement would fall under a block exemption then there would be no need for the businesses concerned to approach the Commission at all. It would only be if there was uncertainty that undertakings would approach the Commission for individual exemption. Nevertheless, the fact that a block exemption existed would not prevent the Commission from stating that particular aspects of individual agreements fell outside the scope of the exemption.

Block exemptions are introduced in the form of Regulations and are granted in respect of specific types of agreements. There are many examples and these have included:

  • A block exemption on exclusive distribution agreements in Regulation 1983/83, replaced by Regulation 2790/99 on vertical restraints.
  • Another on exclusive purchasing agreements in Regulation 1984/83, also later replaced by Regulation 2790/99 – in fact, Regulation 2790/99 (which came into force in June 2000) replaced exemptions on all vertical agreements with the exception of certain serious restraints.
  • One also on patent licensing in Regulation 2349/84, later replaced by an exemption on technology transfer in Regulation 240/96.
  • One on motor vehicle distribution in Regulation 123/85.
  • Another on know-how licensing in Regulation 556/89.
  • Another recent block exemption is that on research and development agreements in Regulation 2658/2000.

Inevitably, the issue of block exemptions is also affected by the introduction of Regulation 1/2003. The new block exemption Regulation 2790/99 creates an exemption for all vertical agreements as a category in their own right. There are limits to its application. There are special rules, for instance in the case of certain sectors such as petrol distribution. It does not apply also to certain types of clause such as non-compete clauses for over five years, those preventing buyers from manufacturing or selling certain goods after the agreement comes to an end, and ‘hard core’ restrictions, such as those covering resale price maintenance or restrictions on resale outside exclusive distribution networks. However, it is wider than the Regulations that it replaced, applying, for instance, to unfinished goods.

The new block exemption is applied according to a market share test. In this way it only applies if the buyer or seller has below 30% of the market share. This does not mean that the agreement cannot gain exemption but an individual exemption will have to be applied for. It is on this basis that the new block exemption has been criticised for lacking the certainty of the previous system.

A flow diagram depicts how a breach of Article one hundred and one is established with ‘yes’ or ‘no’ conditions.

Long Description for Figure 16.1

The flow begins with the question, “Are the bodies concerned undertakings?”. If ‘yes,’ then “Is there an agreement or decision of a trade association or concerted practice?”. If ‘yes,’ then “Does it have the object or effect of preventing, restricting, or distorting competition?”. If ‘yes,’ then “Will it affect inter-state trade directly, indirectly, actually, or potentially?”. If ‘yes,’ then “Is the agreement of minor importance or can it be exempt?”. If ‘no,’ then “The agreement breaches Article one hundred and one”. If ‘yes’ and all other ‘no’ conditions lead to “There is no breach of Article one hundred and one”.

Figure 16.1 How a breach of Article 101 is established

16.3 Article 102 and abuse of a dominant position

16.3.1 The concept of abuse of a dominant position

As we have seen, Article 101 concerns some form of collusive behaviour between undertakings that would have an effect on intra-Community trade and that in any case has as its object or effect the prevention, distortion or restriction of competition. Article 102, on the other hand, is usually concerned with the actions of a single undertaking rather than a combination of undertakings.

On this basis Article 102 was introduced to combat the perceived threat to competition posed by the negative effects of large concentrations of economic power in the hands of individual undertakings. It was also traditionally felt that Article 102 could not apply to oligopolies (large concentrations of economic power in the hands of a small number of undertakings). However, this reasoning has been challenged.

Article 102 is also applicable in the case of mergers and concentrations.

It must be remembered that Articles 101 and 102 are merely different mechanisms for dealing with what is essentially the same problem: the anti-competitive practices. The two Articles are, therefore, not mutually exclusive. Either may be alleged in a specific case and there is some overlap between the two. As a result, the Commission has a discretion how to apply competition law and which Article is in fact breached.

However, while there are overlaps with Article 101, it must also be remembered that there is no possibility of negative clearance in the case of Article 102. Neither are exemptions available (though they are for mergers). Once a breach has been proved then there is no further way of avoiding the consequences of the breach as there is in Article 101.

The basic prohibition is again spelled out in the Article itself. According to Article 102:

Before moving on to analyse the separate elements of the Article it is important to emphasise that Article 102 does not in any way prohibit the existence of dominant positions. It is plain that the Article is not meant to punish efficient economic behaviour. Its purpose is rather to discourage practices that would cause damage to other undertakings by creating artificial conditions that distort competition. In this way it is the abuse of dominance that is significant more than the dominant position itself.

Looking at the wording of the Article, there are three distinct requirements all of which must be proved in order for there to be a breach of Article 102:

  • The undertaking engaging in the practice complained of must have a dominant position in the appropriate market (which can be either the Single Market or a substantial part of it).
  • The practice in question must amount to an abuse of that dominant position.
  • Trade between Member States is affected as a result of the practice.

16.3.2 The concept of ‘undertaking’

Articles 101 and 102 are both concerned with the anti-competitive activities of ‘undertakings’. The definition given to the term ‘undertaking’ then is the same as under Article 101 (see section 16.2.2).

The term is thus defined broadly. Again, it is not defined anywhere in the Treaty, so again we have to look to the case law for interpretation of the term by the ECJ: ‘a single organisation of personal, tangible and intangible elements, attached to an autonomous legal entity and pursuing a long term economic aim’.

Again, it covers almost every type of entity regardless of its legal status, from an individual to a multi-national corporation, provided that it has legal capacity and is engaged in an economic activity. However, as with Article 101, the definition is capable of evolving on a case-by-case basis.

16.3.3 The definition of ‘dominance’

‘Dominance’ is not defined in the Treaty at any point. However, it was originally defined in the ECSC Treaty (European Coal and Steel Community Treaty). The definition given here was that dominance occurs where undertakings hold a position ‘shielding them against effective competition in a substantial part of the common market’.

This is a fairly limited and imprecise definition and so as with other significant terms it has been left to the ECJ to define in the case law. The earliest possibility came in Continental Can Co v Commission (Case 6/72) [1973] ECR 215 where the ECJ identified that dominance amounted to:

Two clear elements arise from the definition in the above case: the share of the market enjoyed by the undertaking and its ability to act independently. It was not long before the ECJ had the opportunity to review, re-affirm and supplement this definition.

The ECJ, building on its definition in Continental Can, identified that dominance is:

This again emphasises the significance of the ability of the undertaking to act independently without regard to competition. Another significant feature was added by this statement of the ECJ, the importance of the relevant market. In other words the ECJ was acknowledging that market share had to be analysed by reference to the specific market in which the undertaking was competing. In the case there was discussion whether the relevant market was fruit or whether there was a specific market for bananas.

Yet another extension to the definition was provided in the case of Hoffmann La Roche v Commission (Case 85/76) [1979] ECR 461:

As a result of this gradual development of a definition it is possible to see that there are two critical concepts in determining dominance:

  • the relevant market (this will be determined by analysing not only the relevant geographical market but the relevant product market also – other considerations include whether there is in fact a temporal market, and also cross-elasticity of supply may be important);
  • calculation of the market share of the undertaking in question.

16.3.4 The relevant market

The relevant product market

Obviously, in terms of defending their position and avoiding a finding of dominance, undertakings will want to argue that the relevant product market should be defined as widely as possible. Inevitably the party complaining that it has been affected by the alleged breach of Article 102 will want the relevant product market to be defined narrowly so there is a greater possibility of dominance being found. Key Points from the case of United Brands v Commission (Case 27/76) [1978] ECR 207, above:

Comments