2. Europeanization

A central aspect relates to what can be understood by the impact of the EU on national welfare state policies. The traditional understanding of the study of the EU and its impact has been revolving around neo-functionalism, neo-federalism, neo-realism, liberal inter-governmentalism, and governance or fusion.Reference Wessels and Monar⁹ The EU has directives and recommendations as direct instruments, but also plays the possible role of agenda-setter. The ability to act as agenda-setter can influence the ability and the choice of how to deliver social services in modern welfare states. Furthermore, the European court of Justice has acted and has had a role influencing the development of national welfare states.Reference Martinsen¹⁰ The viewpoints of the commission also influence nation states, especially when the commission argues that it will open a case against a country. We also note that economic policies seem to have had the strongest impact on European integration.Reference Barbier and Colomb¹¹

An analysis of European integration and an understanding of its mechanism is not easy given that Europeanization is not a specific theory, but more ‘a way of orchestrating existing concepts and … contribut[ing] to cumulative research in political science’.Reference Radelli¹² Furthermore, there are at least three different ways to understand Europeanization.

1. (1) The development of a European level of governance.

2. (2) The process of impact and influence from the supranational level to the national level.

3. (3) Processes focusing on the development of shared beliefs and norms that might then be transferred to nation states, but also in both a bottom-up and top-down understanding.Reference Börzel and Risse¹³

However, in general, Europeanization can be understood as the impact of the EU on European integration and on member states, in line with Radelli,Reference Radaelli¹⁴ as this implies a focus on the link between the supranational development and national welfare state development. The impact can, in principle, be in both directions, based on either uploading or downloading of ideas and policies, but in this article the focus is on the possible top-down impact.

There are several approaches to the understanding of Europeanization. One is to see that ‘Subjective Europeanization refers to Europe’s growing role in the cognitive, affective and normative perceptions and orientation of people, and the weakening of the fixation on the nation state. Europe appears as an additional frame of reference, superimposed on the level of the nation state but without necessarily replacing it’.Reference Mau and Verwiebe¹⁵ This interpretation would imply only limited impact on welfare in the member states, although as we will see later this in fact depends on the type of welfare state.

The impact of Europeanization on a national welfare state can be direct, by a duty to impose a directive and/or change due to a court decision. In principle, the impact can also happen indirectly through the gradually more integrated European economies. A core argument is that ‘most of the effects of economic and political integration within the EU: on social insurance and social services appear to have been indirect, with high unemployment serving as the main catalyst’.Reference Korpi¹⁶ A distinction has to be made between positive and negative integration.Reference Kuhnle¹⁷ It has been argued that EU social policy is driven mainly by negative integration, or by courts and markets. Furthermore, that positive integration is not done by the transfer of sovereignty from national states to the EU level, but more through collaboration among the member states.Reference Kvist and Saari¹⁸ Europeanization thus is also argued to be something that ‘walks on two legs – market efficiency and political voluntarism’.Reference Verdier and Breen¹⁹ I focus here on negative integration as a consequence of external pressure from the EU on national welfare policies through regulation and court cases.

In principle, the EU might influence national welfare policies not only through ‘hard’ law (e.g. such as directives), but also steering it through the OMC (Open Method of Cooperation). This is not my focus here, and given the difficulty with a counter-factual position it is difficult to present evidence thereof.Reference Falkner²⁰ A simple model of the Europeanization of social protection thus looks into how developments at the EU level regarding policy processes, the internal market and so on, are mediated by the welfare regime, country size, etc – this then might imply different kinds of responses and welfare reforms. The conclusion just a few years ago was that since 2000 the EU has had an increased influence on social protection, but also that ‘in sum, EU developments have influenced national welfare reform, but in varied ways and to different extents’ (Ref. Reference Kvist and Saari18, p. 238).

The impact of the EU has often been measured by looking into whether or not convergence or divergence takes place. Convergence can be related to factors such as ideas, structure and spending; it might be equally argued that it is easier to agree upon ideas than to implement them,Reference Radaelli²¹ which indirectly also can be seen to be confirmed by the fact that a common understanding of the concept of ‘flexi-curity’ and policy aims of transforming the labour market policy from passive to active are in general accepted by member states without looking into the details of a more precise understanding and implementation of an active labour market policy. Convergence in the labour market policy seems to have taken place.Reference Aurich⁶ However, it can be that this is in the instruments used, and further, that convergence might not always be the same as Europeanization.Reference van Vliet²² A further possibility is that it is the level of social benefits that is converging, although evidence is mixed given that unemployment benefits seemingly have converged whereas this is not the case for social assistance.Reference Caminada, Goudswaard and van Vliet²³ Convergence on a lower level of benefits opens a race to the bottom.

Convergence has been one of the expected outcomes of a more common approach and of the integration of the European economies, also revolving around issues such as policy innovation, diffusion and transfer. There have therefore also been several approaches to try to find out whether there has been an impact from analysis using different dependent variables such as goals, content, instruments, outcomes and style of policy-making, also with the argument that policy output might be correlated with, in particular, the levels of technological and economic development. From 1985 to 1999, for example, there seems to have been a trend towards convergence within the EU.Reference Alsasua, Bilbao-Ubillos and Olaskoaga²⁴ An analysis of the EU 14 (excluding Luxembourg) in the time-span from 1980 to 2005 showed that in central welfare issues such as pension and unemployment there has been a tendency towards convergence,Reference Paetzold²⁵ in the classical way of catching up. In the same vein, there has been an attempt to analyse whether there is convergence in convergence research.Reference Heichel, Pape and Sommerer²⁶ It has also been pointed out that OMC might ensure an Europeanization of the employed working in the different welfare states in various policy areas.Reference Teague²⁷ However, it is also obvious that the capacity of convergence policy is lower within the OMC given the focus on bench-marking, recommendations and peer-pressure compared with situations where direct legally binding regulations or court decisions are available.Reference Citi and Rhodes²⁸ Still, the indication is that there have been pressures from the EU towards integration also having an impact on national welfare systems, and not only in relation to benefits, but also services.Reference Sindbjerg²⁹

According to Falkner,Reference Falkner²⁰ then, it seems that the EU overall has developed into a multilevel, but highly fragmented policy system, and also a system with shared competences. FalknerReference Falkner²⁰ also argues that one form of EU social policy is coordination in order to stimulate voluntary harmonization in the core fields of employment policy, social assistance, pensions and education. Here the central focus is that the rules supporting coordination also imply that a specific welfare state type with universal access to benefits is increasingly difficult to maintain.

3. Core Areas in the EU in Relation to the Welfare State

Historically, the EU has in principle had only limited influence on the structure, financing and type of the national welfare states. The increasing demand for fiscal prudence, especially within the EMU, can influence the options available for nation states. The possible impact stemming from the EU’s understanding of, and use of, the European structural fund, and pressure for taking specific actions in relation to it also have an impact (for a possible scenario cf. Ref. Reference Verschraegen, Vanhercke and Veerporten30 for the impact on activation in Belgium).

From the early days of the EU until the present, the core influence of the EU has been related to the free movement of workers. Rules were laid down with the now changed Directive 1408/71 (Regulation 883/2004) as the core instrument for ensuring the easy free movement of workers. The regulation originally covered employed and self-employed people, but was later broadened with Regulation 883/2004 so that all those ‘who are or have been subject to the legislation of one or more Member States’ are covered.Reference Pennings³¹ In combination with the case law of the European court and combined with the increased internal migration especially in the wake of EU enlargement from 2004 onwards, and the fiscal crisis since 2008, this has increased the possible pressure on universal welfare states.

The general idea behind free movement was that if a worker moved from one country to another, the rules should not be a hindrance, but instead should support that free movement. For example, this should make it possible to combine the earned right to pensions from different countries; the right to welfare benefits should also be given to workers; child benefits should be available even if the child is not residing in the country where the work takes place. Rules preventing people with disabilities have also been said to hinder free mobility, given the restrictions on exporting benefits and non-legal barriers in several areas such as transport and housing.Reference Morgan³²

The use of regulations and other attempts to harmonize has been the main EU approach for a long time, until it became clear in the mid-1990s that because most areas needed unanimity, it would be difficult to decide on binding directives. There have, for example, been directives on social security (1997/7/EEC) and equal treatment (1976/207/EEC) that could have an impact on the nation state’s ability to pursue welfare policies. Later there was also a directive with regard to leave in relation to the birth of a child (Parental leave 1996/34). Still, in general, the EU as such has not been a central player in having a direct impact on nation states’ welfare policies, with the court of Justice presumably the exception. Nevertheless, the indirect influence due to free movement can be noticed (cf. below on the Laval case and the debate on the service directive). The fiscal crisis, leading to a demand to fulfil the EMU criteria if receiving support from the EU, which in several countries includes a reduction of the public welfare sector and its structure, also provides an example of the growing impact of the EU on welfare state policies.

Since the Amsterdam Treaty, and with a better focus on how to promote employment, there has also been a concern with regard to the reconciliation of work and family life. This has included a focus on gender equality policy, equal treatment at work, parental leave, and child care. There have been directives on equal opportunities, on services in the internal market and on equal treatment (also one of the cornerstones of the Rome Treaty), communication on a better work–life balance, council conclusions and targets related to the employment rate of women, etc, and also increased funding (European Social Fund) in these areas.Reference Jacquot, Ledoux and Palier³³ Thus, in spite of the fact that the EU in general only has a weak role with regard to welfare services, the interaction between EU policy and national policy has gradually become stronger, and case law has moved the boundaries between the ability to make decisions between the EU and the nation states.

The difficulty in reaching consensus also helps explain why the development in the EU and its member states in several areas is often more driven by court decisions than by EU formal decisions, for example in the area of health care services.Reference Greer³⁴ The court’s starting point in relation to welfare issues has been an interpretation of the aim of the treaty including a gradually higher level of integration and a reduction in direct or indirect hindrances to free movement. The implication is that the EU has been more focused on the use of the market and thereby on liberalization than on making formal decisions about how and to what extent there should be an overall way of regulating welfare state developments.

For example, part of the reason why there have been court decisions in the area of health care is that it has been difficult to draw a line between hospital and non-hospital care services. This presumably also explains why health care was taken out of the service directive. Implicitly, this meant that the court was still left with the free movement of goods and services as the main guiding line when regulating health care. The distinction made differed based upon whether the person needing treatment has to be hospitalized for more than one day. The implication is that day-to-day services can be treated as part of the free movement of services, thereby reducing a nation state’s option to regulate the area as part of a welfare approach. In other words: ‘there is essentially no evidence of economic, political, or social supporting coalitions behind the ECJ[European Court of Justice]’s move into health care services’ (Ref. Reference Greer34, p. 191).

It is with this focus on how European integration influences national welfare states, especially through free movement, that the following section looks into the impact on a universal welfare state: Denmark.

4. Changes in Welfare Approach in Denmark as a Consequence of Free Movement

In a universal welfare state, as is Denmark, there can be important implications if the basic criteria for receiving benefits as a citizen or a person legally living in that state are undermined by EU regulations or court decisions. Such undermining can be related to the welfare state’s legitimacy, the financing of the welfare state, or the inner cohesion of the welfare state. The focus here is on whether free movement and the rules and understanding thereof have de facto implied changes to the Danish welfare state, or changes as a consequence either of the understanding of social rights, due to recent court cases, or viewpoints of the commission. Even though the case here is specifically about Denmark, it will be argued that changes are in principle also far reaching for other countries having social security systems based on universality either in part of as a whole. The analysis will be based upon official documents, and will cover the central income transfer areas, with a special focus on recent changes.

When Denmark entered the European Union it was necessary to change the fully universal access to the state basic pension system, where the only rule for receiving pensions was reaching the pension age, towards a system with rights to a pension depending on the number of years of having lived in Denmark, with 40 years in Denmark, as in many other Nordic countries, now needed to obtain the maximum state pension. In order to receive a full disability pension it is now also a requirement that after the age of 15 one has lived most of the time in Denmark. Otherwise one will receive only proportionate pension benefits.³⁵ Disability and old age pensions are, if one is eligible for them, transferable if one starts to live in another country. Thus, the universality of the pension system was already reduced when Denmark entered the European Union.

There is no required length of time one has to have lived in Denmark in order to receive sickness benefits. This presumably is due to the fact that the first 21 one days of sick leave are paid by the employer where one has been employed for at least 8 weeks with a minimum of 74 hours of work. To receive state sickness benefit a total of 13 weeks’ participation in the labour market with at least 120 hours of work should be documented. Thus, this seems to be less of a risk for the universal welfare state, although, as also argued in the report,³⁵ it is difficult to control, for example, whether sickness benefits should still legitimately be paid out when a person after becoming sick has moved to another country. In 2009, 10,600 persons from the EU received sickness benefits with a total cost of approximately 46 million Danish Kroner. The right to maternity and paternity leave is dependent on 13 weeks of employment, and here work done in another country in the EU can be included in the calculation, if migrants are legally residing in Denmark.

The rules related to social assistance stipulate that one should have lived seven out of the last eight years in Denmark and during that period have been fully employed for at least 2.5 years. It is argued that social assistance is a social advantage that, unlike other social benefits, cannot be taken away from Denmark – and this also applies to EU citizens. However, whether this relatively long timeframe can be upheld is questionable, as we will see later.

Early retirement benefits at first could not be exported to another EU-country. This was later changed under pressure from the EU, and given that this system also requires long-time membership and specific payments into an early-retirement benefit scheme, one can question whether this is a universal scheme under threat from the EU.

If one has worked and been a member of an unemployment insurance fund in one’s home-country for 9 months, one can – after three months’ work in Denmark – receive the relatively higher level of unemployment benefits in Denmark. In 2012, 11,000 foreign citizens got Danish unemployment benefits. The portable rights as part of the free movement may thus put pressure on the Danish unemployment system.

The Danish welfare system is also impacted by cases in which Denmark is not directly involved, with the by now classic example being the Laval judgement of 2007, which implied a break with the Danish labour market model. Consequently, Parliament in 2008 created legislation in areas that had previously been taken care of by the labour market partners, introducing a rule on how to define minimum wages under the law concerning the posting of workers.Reference Kristiansen³⁶ Kristiansen also argues that the EU rules concerning the free movement of workers seen from the employee organisation point of view generally imply the possibility, and therefore the risk, of social dumping. Thus, the rules of free movement also increase the risk for other parts of the welfare state or for its labour market model.

In 2002, the rule on how long one had to have lived in Denmark in order to be eligible to receive social assistance was introduced. People who had only recently entered the country would only be eligible for so-called start-help. Start-help was abolished on 1 January 2012, but its one-time existence indicates that the requirement on time one has to have lived in Denmark has been used to reduce the number of persons that get access to social assistance, and under which conditions, and to relieve the pressure on welfare state spending, But it thereby also did away with universality in the welfare state. However, as will be discussed later, it is an open question how long a time spell required to acquire social rights will be accepted by the European court. Thus, the rules on social assistance may be in need of revision, and less time in Denmark will be required before being fully eligible for social assistance.

Similar changes in the family allowance system were enacted in 2010. Full child benefits would first be received only when having lived in Denmark for at least two years. After 6 months either in a row or within the last 10 years 25% would be paid, and this would gradually increase so that after 2 years within the last 10 years full benefits would be paid. In the proposal for the change³⁷ it was argued that the main reason for the change was that child and family allowances fall within the scope of 883/04 and that therefore without the change persons entering Denmark covered by labour market regulations would immediately be eligible for full child benefits even when the child was not residing in Denmark. So, in order to avoid that persons working in Denmark for a short spell should have the right to full benefits, the law was changed. In the proposal to Parliament it was also argued that people ‘should have a real connection to Danish society to be able to receive full benefits’ (author’s translation). In November 2012 a slight change was made to ensure that refugees with children not fulfilling the above mentioned requirement could receive full benefit as of day one.

However, in June 2013 the Danish Government announced that: ‘It is our conclusion, that it is not possible that workers from other EU-countries should have a certain amount of employment, before they can receive benefits’.³⁸ This conclusion was reached following the EU-commission arguing that such restrictions are against the rules regarding free movement concerning coordination of social security, including the aggregation principle.³⁹ To be considered as being employed in Denmark one should work at least 9 hours per week, and to be counted as self-employed the requirement is at least 18.5 hours.

The number of people from Eastern Europe with a right to family allowances in Denmark increased from 2676 in 2006 to 5772 in 2010, and from the EU in total from around 11,000 in 1995 to 19,000 in 2010.⁴⁰ This increase in the numbers and the lack of options to reduce the right to get benefits has opened up a public and political debate on the legitimacy of the welfare state in the sense that some voters are not prepared to pay benefits to children not living in Denmark, and therefore, next to the increase in spending, also to the questioning of the legitimacy of the universal Danish welfare state.

In addition to the changes already mentioned, in 2010 a committee was installed to look into the possibility of enacting in more areas that one should have resided in Denmark for at least some time to be eligible for various welfare benefits. A report was published in March 2011.³⁵ The report focused on the challenges of migration, including free movement, in relation to welfare benefits and, further, in certain social services areas, including free education. In the area of educational grants recent cases also illustrate the risks on spending and access for a universal welfare system. It is not only with regard to benefits that it was argued that there might be a need for waiting periods.

A court decision of 21 February 2013⁴¹ implied that if a person takes up a job as a migrant worker access to educational study grants has to be granted. The court argued that free movement: ‘must be interpreted as meaning that a European Union citizen who pursues a course of studies in a host Member State whilst at the same time pursuing effective and genuine employment activities such as to confer on him the status of “worker” within the meaning of Article 45 TFEU may not be refused maintenance aid for studies which is granted to the nationals of that Member State’ (Case C-46/12 – ruling on the 21 February 2013). The cost to the Danish welfare state has been estimated at around €30 million, and has raised the issue about how in this area also to require that one has lived in Denmark for some time before being eligible for study grants.

In another case (Giersch – C-20/12) a preliminary ruling stated that for frontier workers Luxembourg could not deny children study grants with the existing criteria related to conditions of stay, as the existing ‘condition nevertheless goes beyond what is necessary in order to obtain the objectives pursued’ (Giersch – C-20/12 p. 14).⁴² There is thus room for some criteria, but presumably not a strong demand of a long-time stay in a country before being eligible for benefits. This is also the opinion of the Advocate General (Joined Cases C-523/11 and C-585/11) regarding funding abroad for higher education. Germany has a rule demanding three years of residence before having such a right, and in the case of two children who had been abroad with their parents, who had been working in other countries, the children were denied benefits. From a proportionality argument three years is too long, although this is not an argument that a certain time condition is not at all possible; however, most likely only a relatively short time up to around one year will be accepted. The Danish educational grant system will thus also need to be changed as Denmark has a two year rule out of the last 10 years. This also raises the issue of the size of the benefits: if the high Danish grants were reduced it might become less interesting to move to Denmark in order to obtain a study grant.

The different cases⁴³ and also historical changes depicted above are a clear indication that the combination of EU-rules on free movement with court cases and Commission intervention exert pressure on the legitimacy and the cost of a universal welfare state such as Denmark. There will thus presumably also be a constant pressure to change these systems mainly in such a way that length of stay, as far as it is still possible, can only have a limited impact on the right to and level of welfare benefits. Other ways will be to reduce the size of the benefits and make discretion more central and by this reduce the pressure on spending. This will reduce the universality related to citizenship, and can move the system towards a more social insurance based system or a more targeted system based upon and within the social assistance system. The changes might not only affect welfare benefits, but also social services, as instanced by the service directive.

5. The Service Directive – A Case and Possible Impact on Social Services

The Service Directive provides a further and also more general case, illustrating how the principles of state and market involvement sometimes interact and can lead to different outcomes, and that if the court especially has a focus on market and market provisions this will drive a wedge in the provisions and the structure of the universal Danish welfare service model and over time lead to less strong state involvement. The central issue related to social services is whether the service is a good that should be provided by the market and be available for purchase on the market at a given price, or whether it is a public good provided as a result of some of the logic for providing public goods. Over time the boundaries between state, market and civil society have changed. Care for the elderly could, as an example, in principle be provided, not only – as earlier – mainly by the family, but also by the market. Given that it is very difficult to estimate the precise need for elderly care before the need actually arises, this has been coped with financially by the welfare state in one way or another, seen, as argued by Barr,Reference Barr⁵¹ that the welfare state is a Piggy Bank. The same to a large degree with health care, where recent years have seen some changes.

Part of the reason for the development of the Service Directive has been the change in case law, or more specifically that a number of cases indicated that ‘judicial activism has clarified that the Community principles of free movement of goods and service apply to the health care policy field’,Reference Martinsen⁴⁴ and that national rules regarding the authorization of who to provide non-hospital care for was a barrier that was not in accordance with the rules of the free movement of services. In general this has implied that since the end of the 1990s a number of European court cases have expanded the citizen’s right to planned health care in countries other than his or her home country.Reference Kumlin⁴⁵ The idea is that people could also travel or use services in neighbouring countries in areas such as physiotherapy, and dental and medical treatment, including visits to a general practitioner and having prescriptions delivered. Nation states supporting this economically will therefore have to pay the same amount to a citizen even if the treatment takes place outside the home country. Individuals also have the right to move to another country to search for a job and claim unemployment benefits in another country. So, the implication is that it is not only free movement of labour, but also free movement of goods and services that poses problems for welfare states and thus not only as discussed in Section 4 to welfare benefits, but also in relation to social services.

In principle, as argued in Section 2, the starting point is that the EU in general only plays a limited role in welfare state policies. At the same time, the court has by referring to the free movement of goods and services changed the rules and the way the system works (cf., for example, the three cases regarding Kohl, Decker and Peerbooms). A good example is the Watt case, from 2006,Reference Martinsen and Vrangbæk⁴⁶ where the court arrived at the conclusion that Mrs Watt could get the cost of a hip-operation reimbursed, in spite of the fact that the latter took place outside the physical borders of the UK. The court also argued that medical services fall within the scope of freedom with regard to the provision of service. The Decker and Kohl cases changed the Danish perception from it being a national issue to the fact that health care was clearly not outside the scope of the internal market. This led to a change in the Danish legislation allowing for the purchase of specialist services (e.g. dental assistance) abroad and then being reimbursed with a fixed amount of money equal to what the Danish providers would get; there was also the interpretation of ‘services as services normally carried out in return for remuneration’ (Danish report – here quoted from Ref. Reference Martinsen and Vrangbæk46, p. 179).

In this way part of the welfare state’s social services are open to international competition, and this also requires a change with regard to steering and governing in these areas. This also includes whether they should be universally available or that more specific criteria are needed to reduce the risk that under the rules on free movement they might easily be used by people who have not been participating in the financing of the universal welfare state.

This became even clearer with the argument over the Service Directive (the so-called Bolkestein Directive). This is a further indication of the possibility of a movement away from the universal welfare state, and of the difficulties it may run into when trying to provide specific social services. If a national welfare state cannot provide appropriate health care services to a citizen without too much delay (and this is not precisely defined), then a citizen may have the right to go to another country to obtain the treatment.

The Directive on services in the internal market is a clear indication of the ongoing debate on, and the struggle between, state and market led developments. The draft directive from 2004 was ‘aimed to reduce national regulation of services, unless they were non-discriminatory, justified with respect to public interest and proportionate’.Reference Grossman and Woll⁴⁷ One debate circled around the country of origin for setting up a business in the sense that those having residence in one country should only abide to the rules in that country. Another issue was related to the inclusion of public services in the proposal for the directive, including services of general interest.Reference Crespy⁴⁸ A third issue revolved around the position of posted workers, where the understanding in the posted worker directive from 1996 was that working conditions and pay should be in accordance with the rules in the country where the work was carried out; the suggested directive was seen as ‘being de facto the ruin of these provisions by making controls impossible’.Reference Crespy and Gajewska⁴⁹ After changes reducing and amending these elements, including the explicit exclusion of services of general interest, the directive was adopted on 12 December 2006.

The court still plays a central role in these areas and it constantly seems to opt against diversity in the structure and rules regarding not only welfare benefits in general, but also the access to social services. This questions the possibility of a continuation of the universal welfare state. This in spite of the fact that, as argued by Pennings,Reference Pennings⁵⁰ if there is a risk of undermining the financial strengths of a social security system this can justify ‘a barrier to the principle of freedom to provide service’ (Ref. Reference Pennings50, p. 435). How strong this risk should be is a still more open question, and the possibility to create barriers seems weakened by the recent court cases.