Ministry of Social Affairs and Health




    The social protection systems of Europe are characterised by a great deal of diversity. They represent a unique product of each nation’s individual historical development. Social protection systems are created according to the relationship between various social and political actors such as the state, regions, localities, families, employers and employees. These relationships in turn are effected by the social, political and even religious climate in which they have had to grow. The distribution of power between these different actors is thus distinct in every nation and entrenched in the national psyche by history and tradition. This means that, even though social protection systems all share the same basic aim of protecting people against social risks and poverty, the methods and terminology used to achieve this vary from one state to another.

    This divergence of terminology has been the subject of a recent survey by the European Institute of Social Security, the results of which were analysed by Professor J. Berghman in his article Concepts in Social Security: The Lack of a Common Language34). Prof. Berghman found that different social provisions mean different things in each state. The survey asked experts in the governmental, academic and administrative sphere to classify a range of social provisions into their respective policy fields. In this case "social provisions" included issues such as unemployment benefits, parental leave, health care, employment promotion or care for the elderly and handicapped, whilst the "policy fields" involved the following classifications, social security, social policy, social protection, employment policy and labour relations.

    34) Social Challenges of the EU and Intergovernmental Conference a publication by the Finnish Ministry of Social Affairs and Health 1996

    Although some general trends could be isolated the results of this extensive survey do indicate that there is some real disagreement about the terms used in the social field. One example is the "container concept" of social policy, which could be interpreted on several levels. A narrow definition of social policy includes only study grants and risk prevention, while a slightly broader interpretation adds care provisions to study grants and prevention. The widest construction of social policy represents an all-embracing concept, covering social provisions related to employment and working provisions. Thus, when legislators, academics, lawyers and judges are discussing social policy at a European level they may all be thinking about entirely different concepts. The survey described above has indicated that an effective social dimension in the EU requires far more than high quality simultaneous translation, it needs some form of common structural understanding.

    In the co-author’s analysis of the same survey35), it is revealed that there is some disagreement between experts as regards the meaning of certain European terms or "Euro-speak". The co-author uses the example of the phrase "social dumping". On closer inspection it appears that there are different interpretations of this commonly used term. Some experts thought that it referred to Member States shedding the social rights of their citizens in order to improve international competitiveness; while others thought that it described the migration of benefit recipients to states where social protection benefits are higher. This latter interpretation is often called "social tourism" rather than social dumping. A final group of experts indicated that social dumping embraces both these interpretations. Thus, not only is there conceptual divergence about how to classify national social systems but there is also disagreement between experts on how to construe European terms.

    35) Social Challenges of the EU and Intergovernmental Conference a publication by the Finnish Ministry of Social Affairs and Health 1996

    Closer inspection reveals that a consistency of terminology is lacking even within the social law of the EC itself. For example the term "worker" means different things in different pieces of primary and secondary EC law. In articles 48 and 51 EC Treaty it clearly means wage earners as opposed to self-employed persons. Whereas in the EC Charter on the Fundamental Social Rights of Workers, the concept of worker covers anyone performing an economic activity, which includes the self-employed.

    The fact that Europe’s legislators, lawyers and judges lack a common language in the field of social protection could inhibit social and economic convergence. A common terminological and methodological structure is lacking here. Perhaps a first step towards rectifying this problem is to make sure that everyone is aware that they are using terms that maybe understood differently from one state to another. The survey conducted by the EISS has provided important information as to how the same term can be given a variety of meanings. The first step is thus to proceed with caution, as Prof. Berghman explains:

      "[The survey] indicates that even among experts unanimity on unequivocal terminology is lacking. This should warn us to be very careful in our reading and writing on the social field. It should warn the policy agents at the European level even more when devising solemn declarations or even legally binding texts.36)"

    36) ibid at page 17

    Some EC legislation fails to define important terms contained within it. Precise definitions will allow legislators at the European level to predict and avoid some of the problems caused by ambiguity and a lack of common terminology. These definitions will allow the national representatives in the Council to predict with increased certainty how the new laws will effect the national social protection system. Legislating with increased certainty will reduce the shocks that some states have received from ECJ interpretations of laws that they had previously consented to in the Council. For example, one could look at the case law of the ECJ, which has been necessitated by the fact that there is no definition of social insurance or social assistance in Social Security Co-ordination Regulation 1408/71 EC. Even though this Regulation was accepted by unanimity, recent ECJ decisions have led some states to question the fundamental structure of their social protection systems. It is thus clear that when they accepted the EC Social Security Co-ordination Regulation, they were not thinking of social assistance in the same way as their fellow legislators and, ultimately, the ECJ.

    Common definitions will also contribute to a clearer assessment of existing legislation. Discussion of the positive and negative effects of EC rules could be substantially hampered if experts are not communicating on the same wavelength. As well as allowing the easier assessment of EC rules, common definitions might also enable easier discussion of national policies and the transmission of ideas. A shared understanding of particular social terms might even lead to some natural convergence of the national systems. This will result from the transmission of best practices over national, linguistic and terminological borders as well as through the gradual implementation of the common definitions into the national systems.


    At the moment social policy in Europe is characterised by a divergence in terminology. Experts are heavily influenced by their own unique national systems and may fail to appreciate that their colleagues in other Member States maybe thinking of something very different when discussing important concepts such as social policy or social dumping. Terms such as social insurance and social assistance have filtered through a unanimous voting procedure without all the states being absolutely certain what they mean. The ECJ has had to step in ex post in order to resolve definition problems that could have been dealt with in the legislation itself. Therefore, a lack of common understanding, which ultimately leads to conflict, has to be determined by the judiciary rather than any (indirectly) democratically elected body politique.

    Comparison between states is complicated by a lack of common definitions and uniform statistics. However, a realisation that social policy lacks a common language is slowly growing within Europe. The EISS survey has revealed that divergence of terminology in the social field clearly exists. This however has just touched the surface and greater research is required to determine the true extent of this problem and what it might mean for the development of a European social dimension.

    Uniform statistics based upon uniform concepts such as unemployment or incapacity are very important in allowing researchers, legislators and the general public to compare their states with others in the EU. The task of compiling such statistics has been given to EUROSTAT or the "Statistical Office of the European Communities". EUROSTAT collects data from each Member State and then attempts to harmonise it using adjustments that take into consideration the special characteristics of every nation. The development of a common statistical language has become one of the express goals of EUROSTAT. This common language is beginning to take form through the development of the European Statistics System (ESS).

    Although mainly used by the Commission, EUROSTAT statistics are available to everyone, including trade unions, politicians, industry, universities, the media and private individuals. The data is available in both hard and electronic forms and can be accessed on the internet.37)

    The development of uniform statistical concepts will be of great assistance in the creation of comparative tools such as social indicators. The EU already places a great deal of reliance on economic indicators which have played an important role during the realisation of the EMU convergence criteria and which will continue to be of vital significance following the introduction of the single currency. The absence of social indicators is another example of how economic ideology has dominated the process of European integration and created a social deficit. Uniform concepts that relate to economic factors such as price stability and interest rates appear frequently within the EC Treaty. The EU could use a standard definition of a term like "unemployment" to monitor the labour markets in the Member States, in the same way that it monitors economic indicators such as public deficit. If the EU then publishes the information it has collected, the people of Europe will receive a clear picture of how well their social protection system or job market is operating in comparison to their European neighbours. It is important to stress that social indicators are not objectives, they merely represent the situation as it already exists and are not goals or targets. Of course the development of uniform and acceptable social indicators can be used to determine whether or not a given Member State has achieved a particular EU target. It could be said that establishing mutually acceptable social indicators forms a pre-requisite to the setting of common EC objectives that is discussed in the next chapter.


    The EU has two options by which to create the common definitions that are needed for increased clarity and certainty in the European social field. The first option would involve harmonising the definitions at national level so that all states are forced to abandon their diversity and accept a common list of terminology. Preparing a definitive list of social terminology would be an extremely complicated and intricate task. Social protection systems can only be harmonised using either article 100 in the existing EC Treaty or through the Social Agreement that shall be added to the EC Treaty after the ratification of the Amsterdam Treaty. Both of these legislative options require unanimous approval in the Council for any measures in the social protection field. This means that the list of common definitions would have to satisfy every single Member State and given the historically entrenched diversity of the national social protection systems, coming to a political compromise on a proposal of this nature will be extremely unlikely.

    The second option involves creating specific EU definitions, which run alongside the national ones. As far as potentially ambiguous terms in legislation are concerned, this can be done with hard or soft law.

    A hard law approach would involve defining all the important terms found in any new legislation within the body of the rules themselves. Careful research should be conducted in order to make sure that there really is a common understanding between the Member States before rules are passed, this research might rely upon comparative studies to isolate terms with particularly divergent interpretations. As regards existing legislation it makes perfect sense that if a legal basis can be used to pass the original norm, it can also be used to define the terms in that norm. Different pieces of legislation could use different definitions, which might be better suited to the task they aim to fulfil, but if possible more harmonised definitions throughout the different areas would be preferred.

    A combination of hard and soft law may even be possible. Each legislative act could be accompanied by a soft law measure, such as a Recommendation or Communication, which defines all the important terms. Although the soft law measure will not be binding in itself it will exercise considerable influence over the interpretation of the ECJ and the national court.38)

    38) see Grimaldi v Fonds des Maladies Professionelles (1991) ECR 4407

    The EU may even elect to create an "EU Social Dictionary" in which those terms that appear frequently in primary and secondary legislation are given a clear and precise definition. These definitions may co-exist with national ones as in the second option described above. The social dictionary would involve one definition for each term and will not allow different meanings to be attributed to different types of social measures. Thus, the term "worker" or "pay" would mean the same whether it relates to co-ordination of social security systems, the equal treatment of men and women or health and safety at work. A dictionary of this kind may be built up gradually through hard law measures. The new Social Chapter could be used for this task, with definitions being built up in each of the respective fields, health and safety, equality between men and women, social security etc, and passed according to the appropriate voting procedure, either unanimity or qualified majority. The disadvantage of this legal basis is that only Directives can be used and these are only binding as to their result. A truly uniform set of EU definitions may require implementation through Regulations. In this case a more appropriate legal basis might be article 235 EC Treaty, the residual legal basis that can be used to achieve EU goals without a clear legal basis elsewhere in the Treaty.

    The social dictionary may also be created through soft law measures. The process of creating such a list of definitions may begin with a Commission White Paper. Recommendations, Opinions and Communications could all be used to clarify certain EU terms.


    The development of a clear set of EU definitions will add greater certainty to the legislative process. This certainty will enable the Member States to predict the effects of new laws before these rules are implemented. Greater certainty will reduce the controversial power of the ECJ, a body that is neither politically accountable nor elected by the people. In the past the ECJ has sent shock waves through national systems with interpretations it has given to legislation, legislation which has already received the unanimous approval in the Council of Ministers. This divergence of opinion between states and court means that the Member State in question misinterpreted the proposal it accepted. The shock waves caused by ECJ decisions create bad publicity, which in turn endanger the popularity and legitimacy of the EU itself.

    However, increasing certainty by ensuring more precise definitions in legislative norms would slow down and complicate the legislative procedure. The trade off is quite clear, certainty about the true ramifications of the legislation or ambiguity (and even deliberate confusion) within that legislation, which has to be resolved at a later date by the ECJ? It should be pointed out here that vagueness in legislation is often attributed to the difficulties caused by unanimous voting. Unanimity usually demands political compromise and this in turn leads to vague terminology.

    Problems may arise if the concurrent definition approach is applied to statistics. The uniform compilation of data represents a major break through because real comparisons can finally be made between states. The information provided will be useful for academics in the comparative field and will actually allow the people of Europe to determine the quality of their own national social protection systems in relation to others within the EU. The availability of uniform data may put a stop to every social security administration in Europe insisting that it provides the highest benefits or costs the most to operate. However, the problem is that if EU statistics operate alongside national ones and there are substantial differences in the results being transmitted to the people, how is this going to effect the legitimacy of the EU? Which statistics will be believed?

    Great pains must be taken in order to ensure that EU definitions are correctly implemented into the national systems and not influenced by national terminology. It must be made absolutely clear that the EU definitions are distinct from the national ones and that EU and national terminology run side by side. Therefore EU terms may be broader or narrower than national ones. For example under national law a self-employed person might be anyone who is registered with the national chamber of commerce, whereas for EC law purposes the definition might include anyone engaged in independent economic activity. In this example there is a good chance that the EU definition will be boarder than the national one, for instance the EU definition contains no conditions relating to a minimum amount of income or working hours. The result is that some people will be entitled to EU rights for self-employed persons even if they are not entitled to national rights for self-employed persons. If EU definitions are allowed to run alongside national ones, there is a chance that European definitions may slowly be absorbed into the national systems. The constitutional principles of the supremacy of EC law, direct effect and indirect effect would demand that national courts apply EU definitions whenever the facts of the case touch upon an issue of EC law. Thus, EU definitions will be regularly applied by judges and lawyers and shall inevitably come under scrutiny from academics. If there is some divergence between the EU definition and the national one, as in the case of the term "self-employed" mentioned above, there is a danger that some people will be self-employed according to EC law but not according to national law. It might be difficult for people and judges to accept that the same person qualifies for rights under EC law but not under national law. The response of the people to this situation may lead to political pressure to bring national definitions more in line with EU ones and this in turn will bring national definitions closer together.

    Whether the creation of common definitions will lead to national convergence or not, more certainty is needed on a European level. Legislation and social debate are hampered by a lack of common language and a natural tendency to focus on national definitions, which vary from state to state. A better comparison of statistics and an improved understanding of the difference in our social language are called for in order to develop an effective social dimension in the EU.

    Although it is unlikely to provide individuals with any judicially enforceable rights a process of standardised terminology and the movement towards common definitions will provide an important first step or foundation for most, if not all, of the pathways mentioned in this paper.


    The following section will give two examples where precise and uniform definitions will benefit the smooth and certain application of EC law. The first example concerns the need for a set definition in order to distinguish social security contributions from taxes. The second example concentrates on the necessity to develop a clear test to determine whether a benefit is social assistance or social insurance.


    The need for clear definitions of what constitutes a tax and what should be classified as a social security contribution stems from the fact that the social security systems of the Member States are co-ordinated by EC law, whilst their taxation systems are not. EC social security co-ordination is ensured through Regulation 1408/71 EC. This Regulation makes certain that migrant workers are not disadvantaged in relation to their social security position because they have chosen to exercise their right to free movement within the EU. Regulation 1408/71 EC provides, inter alia, that only one state shall be competent for the collection of social security contributions and the payment of benefits. This competence generally belongs to the state where the person is employed39). Therefore, no migrant worker should be obliged to pay social security contributions to more than one state, this was more likely in cases where the individual resides in one state but works in another. Taxation on the other hand is not co-ordinated at an EU level and states are free to establish their own systems. Thus, double taxation is not prevented by EC legislation in the same way that double social security contributions are avoided but it is ensured through bilateral taxation treaties.

    39) There are exceptions to this general rule, the most notable being that of posting, which is provided for under article 14 Regulation 1408/71 EC, see Schoukens, P. (eds) Prospect of Social Security Co-orination published by ACCO 1997

    It is at this point that problems arise with the definition of what is or is not a social security contribution. Some states have been able to take advantage of the ambiguity in this field, particularly when they introduce new charges that are intended to finance social security funds. In general people are more tolerant of social security contributions than they are of increases in income taxes, this is because they get a clearer view of how the money will be spent and how they will eventually benefit from it. Therefore, it is better for a state to tell its people that the charges they are paying are social security contributions. However, it is better for the state to tell the EU that the charge is actually a tax. As taxes are not co-ordinated at an EU level this means that the state can levy the tax on all residents, whether they are socially insured in another state or not. The result is that someone who lives in one state but works in another and is exclusively socially insured in his state of employment will not have to pay social security contributions in his resident state but might still have to pay taxes there. By calling a charge used to finance social security a tax for EU purposes, the state can still charge residents who are socially insured in another country and not violate the EC rules on social security co-ordination. By calling the same charge a contribution for domestic purposes the state benefits from increased domestic political acceptance. The ultimate loser in this situation is the person who has to pay for social security benefits, which s/he will not receive and this is a real disincentive to exercise one’s rights of free movement.

    A new charge has been introduced in France, the General Social Contribution, which the French state has made clear to its people is used for social security purposes but at the same time France insists that this is actually a tax rather than a contribution. A similar tax has been introduced in Portugal and discussion about a greater reliance upon taxation is taking place in Ireland and Germany. Taxation is playing a growing role in many of the nations that have traditionally relied upon contributions to fund their social security benefits. The burden of paying for social security is being gradually lifted from employers and taken over by taxpayers. This is intended to improve international competitiveness by reducing non-wage labour costs. It is interesting to note that states, which have typically relied upon taxation to finance social security benefits, are introducing new social security contributions, such as the Labour Market Funds in Denmark. The result is that the tax/contribution mix of social security financing in Europe is going to become increasingly complicated.

    A clear and precise definition of the borderline between taxation and social security is required. The definition of contribution must be wide enough to ensure that people are not forced to pay for social protection that they will not actually receive. If social security contributions were defined as all payments made from income that are intended to finance social security funds this would prevent states from discouraging free movement by imposing "taxes", which should really be classified as contributions under Regulation 1408/71 EC.


    The importance of developing a clear distinction between social insurance and social assistance lies in the need to establish the material scope of the Co-ordination Regulation 1408/71 EC. In article 4(4) the Co-ordination Regulation expressly excludes social and medical assistance from any of the obligations contained within it. This means that social assistance schemes do not have to be exported to other Member States and they are not subject to co-ordination principles such as aggregation or equality of treatment. No definition of what constitutes a social assistance scheme is provided anywhere in the Regulation. The term has, as could have been anticipated, become of pivotal importance in the application of the co-ordination rules.

    The ECJ has viewed article 4(4) as an exception to a general principle of EC law and therefore given it a very restricted interpretation. The ECJ has made it quite clear that national definitions of what constitutes a social assistance scheme are not relevant in relation to Regulation 1408/71 EC40). The result is that some schemes, which are considered under national law to be social assistance, have been exposed to the EC co-ordination rules. This means that some states obviously did not anticipate the application of the Regulation to these schemes when they agreed upon the co-ordination mechanism in the Council.

    40) Mazzier (1974) ECR 1251

    The jurisprudence of the ECJ has not succeeded in producing a clear borderline between social assistance and social insurance. It has looked at factors such as whether the benefits are based upon need, whether they are related to periods of employment, whether they form a basic or supplementary income, whether they involve a means test and whether they are based upon administrative discretion or subjective rights. The ECJ has even admitted that:

      "[T]he possibility cannot be excluded that some laws fall within both these categories [social insurance and social assistance], because of the class of person to whom they apply, of their objectives, and of their manner of application, and are thus not amenable to any classification.41)"

    41) Callemeyn (1974) ECR 553 at p age 561

    This situation has caused a great deal of uncertainty amongst the Member States which are unable to predict whether social security co-ordination rules can be applied to their new schemes, such as the care schemes recently established in Germany, Austria and Luxembourg, as well as their existing ones. There is some speculation in Germany about whether Regulation 1408/71 EC can be applied to some of the benefits contained in the Federal Social Assistance Act.

    Another problem is that the traditional distinctions between social insurance and social assistance are melting away. Traditional social insurance schemes, based upon periods of employment, have started to rely more and more heavily upon means tests. Whilst social assistance schemes are beginning to provide subjective rights to benefits rather than the classic method of assistance, which relied upon administrative discretion.