ABSTRACT
The article first describes how the European Union has replaced the idea of nationally bounded freedom to move, settle, and work with the idea of free movement for all Europeans and the notion of non-discrimination. The second section gives an analysis of the extent to which citizens of different European countries support the idea of non-discrimination between nationals and European foreigners. The empirical basis for our analysis is the ‘European Values Study’. The descriptive findings show that the idea of non-discrimination is not supported by the majority of the European citizens, and that there are substantial differences between the countries. In the third section, we explain these differences by referring, among other factors, to the level of modernization of a country, the value orientation of the respondent, the level of education of the respondent, and the unemployment rate.
Introduction
Today's European Union began with the creation of a common market for the coal and steel industry. Gradually, other areas and policy fields were incorporated into the process of integration: a customs union was created; a common market and monetary union were formed; and, finally, some EU countries established a common currency regime. This broadening of European cooperation in multiple sectors corresponds to the expansion of European institutions, which are increasingly assuming more responsibilities and gaining greater independence. In fact, EU member states have ceded part of their national sovereignty to the EU. In that EU law supersedes national law, member states and their citizens are directly subject to the decisions made by the EU. The European Commission oversees the implementation of its decisions, and the European Court of Justice has the ability to sanction member states who do not fully comply (Lepsius 1990). A number of indicators illustrate how sovereignty rights have been conferred from the member states to central EU institutions. For instance, the number of decisions made by the European Council and the European Commission is continuously rising (Wessels 1997), the number of Councils of Ministers is increasing (Knill 2001), and intermediary organizations and interest groups are evermore directing their attention to the European level (Stone Sweet et al. 2001; Fligstein and Stone Sweet 2002).
The fact that sovereignty rights have been conferred from the member states to central EU institutions gives these institutions the power to directly intervene in member states; therefore, citizens in member states have increasingly become subjects of EU decisions. It is an open question as to whether the citizens of the European Union accept the decisions and policies of the EU. The citizens’ acceptance and support of EU regulations is significant, especially in determining the legitimacy of European policies due to the fact that democracies are structurally dependent on the support of their citizens (Gerhards 2007). If this support is missing, legitimacy problems can arise for the institutions themselves. This was revealed in May 2005 when the population of France and The Netherlands rejected the European Constitution; the elite project of giving Europe a new constitution failed, as citizens in two member states refused to support the idea. Since then, the EU has been facing a major crisis of legitimacy.
In this article we will tackle this general question of citizens’ acceptance and support of EU regulations by focusing on a specific policy field – freedom of labor (for other policy fields see Gerhards and Hölscher 2003; Gerhards and Lengfeld 2006). The institutionalization of a European common market is a central element in the European integration process, and the idea of free movement of personal and labor is a key dimension of a common market. This freedom enables citizens of one member state to move to another, either alone or with their families, and to work there (permanently or temporarily). The idea behind EU legislation in this field is that citizens from other member states should be treated equally to nationals – EU foreigners should not be discriminated against. In the first section, by interpreting European legislation and EU policies, we describe how the EU has replaced the nation-state concept of free movement of labor and non-discrimination with the transnational idea of equal access to all labor markets for all European citizens.
We then focus on the following question: to what extent do citizens of different nation-states support the notion that citizens from other European countries should enjoy the same rights and opportunities in their domestic job market as themselves. In the second section, we analyze the extent to which citizens support the idea of non-discrimination between nationals and other Europeans in the labor market. The empirical basis for this reconstruction of the citizens’ value orientation is a secondary analysis of the ‘European Values Study’, a representative survey conducted in EU member states and candidate countries.
The descriptive findings show that the majority of European citizens do not support the idea that citizens from other European countries should enjoy the same rights and opportunities in the job market as nationals, and also show substantial differences among countries and individuals. The third section attempts to explain these differences. We formulate several hypotheses, which are tested with logistic regression models.
2 The European Union's idea of freedom of labor and non-discrimination
European societies of the nineteenth and twentieth century are normally characterized as nation-state societies (Gellner 1983; Anderson 1991; Hobsbawm 1992). The main characteristics of the nation-state are the extension of power over all internal affairs as well as the monopolization of force, which aims to stabilize national borders. Military border control, customs, and immigration policy form important fields that demarcate the nation-state from the outside world. The internal territory of a nation-state is structured and stabilized by the police and through the creation of a domestic institutional structure. These institutions range from education, social security, health care, and political systems to the make-up of the national economy.
In addition to the creation of national institutions, nation-states are characterized by the active inclusion of people who live within the national society into their nation-state. These people become citizens of their state, and consequently have the right to elect their government. National citizens can enjoy the services of the welfare system of their country, and only of their country (Marshall 1949/1983; Brubaker 1990). Equal chances are granted to all citizens living within the nation-state. As a result, non-members (i.e., members of other nation-states), are excluded and treated unequally, with the exception of universal and codified human rights that apply to all humans. Foreigners do not inherently have the natural right to settle in another country, to work or to get an education there, to take part in national elections, or to participate in the welfare state.
The process of European integration has tremendously changed the idea of the nation-state, and thus, national institutions. The idea of nationally bounded freedom to move, settle, study, and work is being Europeanized and replaced by an idea in which all citizens of Europe are regarded as equals, who can therefore move, settle, study, and work in any EU member state (Mau 2005; Delhey and Kohler 2006). The borders of the legitimized freedom to move and of legitimized unequal treatment have shifted outwards from national borders to the borders of the EU (Hartley 2003).
The European common market was created by the Single European Act (SEA). The Act was signed in 1986 and is in force from July 1987. The SEA's main effect was to set a deadline for the creation of a fully integrated market by 1992. The common market is defined as an area without frontiers in which free movement of goods, services, people, and capital is ensured. The crucial regulation for our research question is the so-called freedom of movement for employees/workers. This rule guarantees that every EU citizen may look for employment in another member state under the same conditions as a resident of that country (Chalmers et al.2006). The freedom of movement rule also applies to self-employed persons. As a consequence of implementing the freedom of movement rule, the idea of European equality has become a legal claim which guarantees equal opportunities on the job market. Freedom of movement does not only apply to workers, but also for people outside of the labor force, like students and pensioners. The mutual recognition of certifications and the transfer of social security rights for people moving across national borders within the EU were also agreed upon (Hartley 2003). At the EU policy level, the idea of nationally bounded equality, which treats national citizens and European foreigners unequally, has been replaced by the idea of European equality. The idea behind EU legislation is that citizens from other member states should be given equal treatment as nationals – EU foreigners should not be discriminated against.
The term ‘European-wide’ merits further specification, in that transition periods applied to some new EU accession countries. Portugal and Spain joined the EU in 1986; however, complete freedom of movement was not implemented until 1993 due to concern that there would be a strong migration movement from the economically weaker new member states to the wealthier old member states (Kvist 2004). Such migration movements did not occur; on average, less than 2 per cent of all workers decided to move from one EU state to another (Werner 2001: 12; Kvist 2004: 307). Similar transition periods were also issued for the 2004 and 2007 accession countries, because some old EU countries, like Germany and Austria, counted on substantial migration movements, although experts in that field do not think these fears to be realistic. ‘In brief, we find that fears of welfare migration are largely unfounded, but that EU-15 member states have acted as if migration would take place’ (Kvist 2004: 303).2
Even though such transition rules exist for new EU countries, they are limited to a relatively short time period, and apply only to certain countries and groups of people. The idea of freedom of labor and non-discrimination for all European citizens will become valid in the entire EU in the short-term future. Furthermore, the freedom of movement rule includes several additional social rights (European Commission 2004), such as entitlement to the same social security and tax benefits as national citizens and equality in housing (i.e., the right to subsidized housing). Regardless of nationality, the employee is entitled to live with his family, and the family is entitled to receive family allowances on the same basis as national citizens. The freedom of movement rule therefore implies the complete coordination of the social security system. Pension and retirement rights, as well as rights to other social services, are to be transferred across national borders. Consequently, the employee's acquired rights must be preserved, even if he settles in another member state; contributions made to social security in different countries are to be summed up, regardless of the member state in which they were paid. This is done so that the employee is always insured and can immediately benefit from his insurance in the new country if necessary. These regulations show that the idea of equality and non-discrimination for all European citizens is not limited to the job market alone.
3 The acceptance of free movement of labor and non-discrimination by EU citizens
The Europeanization of labor markets brings with it a fundamental reframing and recoding of the idea of equality; this process can be interpreted as a fundamental intervention into the traditional nationally limited coding of equality. But do citizens of the European Union support the EU's regulations regarding the free movement of labor, or do they discriminate between nationals and (European) foreigners? In democratic countries, the concordance of policies with citizens’ preferences is paramount, in that policy is structurally dependent on the citizens’ approval.
We analyze the value orientations of citizens through a secondary analysis of the European Values Survey from 1999/2000.3 The national samples are representative, with at least 1,000 interviewees for each country. The interviewees are over the age of eighteen and were interviewed in face to face interviews. In our analysis, we differentiate between four groups of countries: the fifteen old EU member states; new member states who have acceded since May 1st, 2004; the two states that became EU members in 2007 (Bulgaria and Romania); and Turkey.
The European Values Survey contains one question which is particularly well-suited to operationalize citizens’ attitudes toward non-discrimination between nationals and other European citizens. The question is formulated as follows: ‘Please, tell me whether you agree with the following statement: If jobs are scarce, employers should give priority to [German] people (or the nationality of the respective country) over immigrants’. The respondent could reply with: ‘I agree’, ‘I don't agree’ and ‘neither/nor’. The question is formulated rather restrictively, in that it refers to the idea of non-discrimination under constrained conditions (i.e., when ‘jobs are scarce’). If the respondent agrees with the statement, this signifies a rejection of the EU's transnational non-discrimination policy. The following graph depicts the percentage of ‘I don't agree’ (acceptance of the EU's policy) for each country.
Figure 1 4 shows that there is not a majority approval of the idea of equal treatment between nationals and foreigners in any of the four country groups; however, the level of rejection varies substantially. Whereas 34 per cent of respondents in old EU member states support the idea of non-discrimination between nationals and foreigners, the rate drops to around 11 per cent in the recently acceded country groups. Turkey, with a 30 per cent support rate, comes close to the EU-15 group.
Attitudes toward non-discrimination between national and foreign people: ‘Employers should not give priority to nationals over immigrants’ (%).4
Attitudes toward non-discrimination between national and foreign people: ‘Employers should not give priority to nationals over immigrants’ (%).4
There are clear differences at the national level as well. The majority of respondents in Sweden, The Netherlands and Denmark support the idea of non-discrimination. In Belgium, Luxembourg and Estonia, almost half of the interviewees support this idea. Estonia is the only new member state to reach the average of the old EU members. At the other end of the spectrum, less than 5 per cent in Lithuania, Poland and Malta expressed support for European-wide equal opportunities.
The indicator used refers to ‘immigrants’ in general, and does not differentiate between European immigrants versus immigrants from outside the EU. It may therefore be possible that respondents associated the term immigrant with people from outside of the EU, in which case, their disagreement would not necessarily go against an inner-European opening of the labor market. Unfortunately the EVS dataset does not allow us to test whether respondents differentiate between European foreigners and people from outside of the EU. Two questions in the European Social Survey, however, make such a distinction: The first question asks people the extent to which they think their country should allow people from poorer versus richer European countries to live in their country. The second question asks people the extent to which they think their country should allow people from poorer versus richer countries outside Europe to live in their country. We conducted a correlation analysis between the different items. Pearson's correlations turned out to be rather high: 0.85 for the correlation between ‘immigrants from poorer countries from outside and within Europe’ and 0.80 for ‘immigrants from richer countries from outside and within Europe’ (P < 0.001).5 This result shows that citizens do not distinguish between European versus extra-European foreigners.6
An item from the Eurobarometer No. 53 (conducted in the year 2000) gives us a second opportunity to test the validity of our indicator.7 This survey posed the question of whether one should admit job-seekers from the following groups of people into the respondent's country: Muslims, people from Eastern Europe, people from crisis areas, people seeking political asylum and people from other EU countries. People supporting the idea of equality answer this question in nearly the same way for all of the aforementioned groups. We built an additive index from the different questions: Cronbach's alpha of this scale is rather high (0.89; see Hölscher 2006). It can therefore be assumed that the EVS provides a reliable indicator with which to measure the idea of non-discrimination between nationals and European foreigners.
We cannot, however, conclude that citizens’ attitudes have remained static over time from a single measurement taken at one point in time; rather, the process term ‘Europeanization’ suggests a change in attitudes. The EVS indicator from 1999/2000 had already been posed in some European countries in 1990 and 1995 as part of the World Values Survey. It is therefore possible to test whether a change in the citizens’ attitudes took place parallel to the opening of the job market in Europe. Table 1 shows the results of our analysis.
| 1990 | 1995/1998 | 2000 | Trend | |
|---|---|---|---|---|
| EU-15 | 29.7% | 35.4% | 33.9% | * |
| Sweden | 56.7% | 74.3% (1996) | 77.8% | + |
| The Netherlands | 62.0% | – | 64.3% | * |
| Denmark | 37.9% | – | 54.3% | + |
| Luxembourg | – | – | 44.9% | |
| Belgium | 26.9% | – | 43.7% | + |
| France | 31.2% | – | 36.1% | * |
| Great Britain | 41.5% | 39.1% (1998) | 31.6% | − |
| West Germany | 29.7% | 46.6% (1997) | 29.5% | − |
| Finland | 17.0% | 18.2% (1996) | 28.5% | + |
| Portugal | 8.1% | – | 24.5% | + |
| Italy | 17.8% | – | 20.3% | * |
| Ireland | 28.5% | – | 20.2% | − |
| Spain | 16.6% | 12.6% (1996) | 19.6% | * |
| Austria | 17.1% | – | 17.1% | * |
| East Germany | 24.9% | 21.8% (1997) | 16.8% | − |
| Greece | – | – | 12.7% | |
| Enlargement I | 7.2% | 19.4% | 11.9% | * |
| Estonia | 3.0% | 40.0% (1996) | 43.9% | * |
| Latria | 4.9% | 41.4% (1996) | 18.3% | − |
| Slovenia | 15.1% | 7.5% (1995) | 12.2% | * |
| Czech Republic | 6.9% | – | 9.8% | * |
| Hungary | 10.9% | – | 6.2% | * |
| Slovakia | 5.2% | – | 5.2% | * |
| Malta | – | – | 4.3% | |
| Lithuania | 2.6% | 3.5% (1996) | 3.7% | * |
| Poland | 4.5% | 3.7% (1996) | 3.7% | * |
| Enlargement II | 12.4% | 6.6% | 10.2% | |
| Bulgaria | 10.5% | 6.6% (1998) | 5.8% | * |
| Rumania | 14.2% | – | 14.6% | * |
| Turkey | 21.5% | 17.2% (1996) | 30.8% |
| 1990 | 1995/1998 | 2000 | Trend | |
|---|---|---|---|---|
| EU-15 | 29.7% | 35.4% | 33.9% | * |
| Sweden | 56.7% | 74.3% (1996) | 77.8% | + |
| The Netherlands | 62.0% | – | 64.3% | * |
| Denmark | 37.9% | – | 54.3% | + |
| Luxembourg | – | – | 44.9% | |
| Belgium | 26.9% | – | 43.7% | + |
| France | 31.2% | – | 36.1% | * |
| Great Britain | 41.5% | 39.1% (1998) | 31.6% | − |
| West Germany | 29.7% | 46.6% (1997) | 29.5% | − |
| Finland | 17.0% | 18.2% (1996) | 28.5% | + |
| Portugal | 8.1% | – | 24.5% | + |
| Italy | 17.8% | – | 20.3% | * |
| Ireland | 28.5% | – | 20.2% | − |
| Spain | 16.6% | 12.6% (1996) | 19.6% | * |
| Austria | 17.1% | – | 17.1% | * |
| East Germany | 24.9% | 21.8% (1997) | 16.8% | − |
| Greece | – | – | 12.7% | |
| Enlargement I | 7.2% | 19.4% | 11.9% | * |
| Estonia | 3.0% | 40.0% (1996) | 43.9% | * |
| Latria | 4.9% | 41.4% (1996) | 18.3% | − |
| Slovenia | 15.1% | 7.5% (1995) | 12.2% | * |
| Czech Republic | 6.9% | – | 9.8% | * |
| Hungary | 10.9% | – | 6.2% | * |
| Slovakia | 5.2% | – | 5.2% | * |
| Malta | – | – | 4.3% | |
| Lithuania | 2.6% | 3.5% (1996) | 3.7% | * |
| Poland | 4.5% | 3.7% (1996) | 3.7% | * |
| Enlargement II | 12.4% | 6.6% | 10.2% | |
| Bulgaria | 10.5% | 6.6% (1998) | 5.8% | * |
| Rumania | 14.2% | – | 14.6% | * |
| Turkey | 21.5% | 17.2% (1996) | 30.8% |
Source: World Values Surveys 1990, 1995/98, 2000.
We used a 5 per cent margin of error so that we can discuss changes in citizens’ attitudes. Turkey seems to increasingly accept the idea of non-discrimination. Bulgaria and Romania show no change, whereas the EU-15 and 2004 accession countries exhibit a shallow U-shaped curve. The results are also ambivalent on the national level. In fourteen countries, no significant change in mentality occurred. Support for a nationally bounded market increased in four countries, and support for greater European equality increased in five countries (Sweden, Finland, Denmark, Portugal and Belgium). The findings show that discrimination between nationals and European foreigners did not decrease overall in the period of time we have analyzed. Institutional changes – implementing ‘freedom of labor’ regulations in 1987 – are not supported by changes in citizens’ mentalities.
4 Explaining non-discrimination attitudes
The descriptive results in the previous section have shown that there are substantial differences between countries. This section first discusses explanatory factors that may influence citizens’ attitudes toward equal treatment of nationals and European foreigners, and then empirically tests whether or not these factors have the expected effects.
1. European countries differ in their degree of economic modernization. Karl Marx was one of the first authors to assume a causal relationship between economic living conditions and peoples’ values, and most modernization theories are based on this central assumption. It would exceed the scope of this analysis to reconstruct modernization theory with all its facets, critics and revisions (see Berger 1996; Inglehart 2001; Knoebl 2003 for overviews). We are uncertain even today as to which factors have contributed to modernization and how to determine the causal relations between them. The modernization process results in a one-time historical growth in the economy and the prosperity of the citizens (Maddison 1995: 21). Regardless of how one explains this growth and developing societal prosperity, there exists substantial concurrence among various theorists that modernized societies can only be described – not explained – by a set of characteristics that altogether form a syndrome (cf. Bell 1973; Norris 2002: 20ff).
As economic prosperity increases through the process of modernization, a change in citizens’ value systems occur. According to the work of Ronald Inglehart (Inglehart 1971, 1997; Inglehart and Norris 2003), when chances to satisfy material needs increase, a shift from materialist to post-materialist values takes place. Materialist values include the following: satisfying economic living conditions, security, national identity, and national exclusion. Postmaterialist values, in contrast, are characterized by the desire for self-fulfillment and participation, internationalism, and the opening of national boundaries. Accordingly, we expect that interviewees from economically less-developed countries express less support for the idea that European foreigners should enjoy the same rights and opportunities in the job market as themselves than will interviewees from more economically modern countries.
The societies examined differ in their degree of economic modernization and social prosperity. We used the Human Development Index (HDI) to measure the degree of a country's economic modernization. The HDI includes three indexes: real GNP per capita, the average level of education, and average life expectancy.8 The data set also contains a way to directly measure materialistic and post-materialistic value orientations, owing to the fact that the survey contains all the items with which to construct the so-called Inglehart index. In addition to the HDI macro variable, we also used the Materialism/Post-materialism Index as a micro variable.9 We proceed from the hypothesis that post-materialists are more likely to support the idea of non-discrimination than materialists.
2. The opening of borders and the dismantling of national protections may lead to a higher degree of competition, particularly in the job market. If foreign workers are perceived as a threat to the respondent's own employment status, it then becomes more likely that the interviewee would support closing the national market. This basic hypothesis was formulated and tested in the early phases of social-psychological prejudice research and in the Realistic Group Conflict Theory (Campbell 1965; Sherif 1966). The authors assume ‘that group conflicts are rational in the sense that groups do have incompatible goals and are in competition for scarce resources’ (Campbell 1965: 287). These theories have been empirically confirmed in numerous historical studies on changes in stereotypes and prejudices in regard to conflicts between groups (e.g., Haslam et al.1992), as well as in studies dealing with interethnic problems caused by immigration (Bonacich 1972, 1979; Olzak 1992; Fuchs et al.1993; Quillian 1995). We assume that people who anticipate that an increase in immigrants will create negative outcomes concerning their own chances on the job market are more likely to oppose European-wide freedom of labor than those who do not think this is the case, or even believe that they could benefit from an immigrant work force.
It is often the case that immigrants from poorer countries to wealthier countries have rudimentary qualifications, and that competition therefore occurs with the nationals having similarly low qualifications. Citizens earning higher salaries may, on the other hand, profit from the increased competition in the lower stratum of the job market as a result of wage competition driving down the prices of goods and services. For example, affluent citizens in Berlin profit from the self-employed Polish craftsmen and cleaning women who moved to Germany after EU enlargement to the East. However, German craftsmen and building workers have an entirely different view of the Polish immigration (Gerhards et al.2007). Due to the influx of cheap labor, competition in the job market increased and wages decreased. For low-skilled workers, the EU regulation of free movement of labor is connected to concrete material disadvantages. They therefore may adopt an attitude that supports the exclusion of foreigners.
We expect unemployed people in countries with an overall high unemployment rate and those with low educational qualifications to oppose the equal treatment of nationals and other European citizens. We have included the following three variables in our analysis: employed/unemployed, level of education, and the national unemployment rate. We expect that the unemployed people from countries with higher unemployment rates and those with a lower level of education are less likely to support non-discrimination between nationals and European foreigners on the job market.10
3. When expressing support for European-wide equality, people may follow their ideological orientations in addition to pursuing their interests. The left/right scheme depicts an abstract ideological grid that citizens use to interpret concrete political topics. Fuchs and Klingemann (1990) have empirically reconstructed the left/right scheme through an investigation of three countries. In their reconstruction, ‘right’ is strongly associated with national identity, nationalism, the conservation of the pre-existing system, and national exclusion. ‘Left’ is associated with equality, solidarity, socialism, and internationalism. We expect that people with a left-wing orientation support equal treatment of nationals and other European citizens, whereas people on the right end of the spectrum are more likely to support a nationally bounded concept of equality.11
In order to test our hypotheses, we calculated three different logistic regressions.12 The first model only includes two macro variables: the Human Development Index and the Unemployment Ratio. A square term of the HDI was added to the logistic regression, due to the fact that earlier analyses have shown that the HDI variable has a non-linear effect on citizens’ attitudes; but the effect was positive and hence confirms our hypotheses. The second model includes all of the aforementioned individual variables. We also included the age of the interviewee as a control variable. The third model considers both macro and micro variables.
As the pseudo-R2 values in Table 2 show, we can explain the attitude toward non-discrimination between nationals and European foreigners very well with our theoretically deduced independent variable. Both the macro factors and the micro variables influence attitudes toward the equality concept of the citizens. In the integrated third model, these factors can explain 25 per cent of the variance.
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| Macro-variables | |||
| HDI | −18.478*** | −19.260*** | |
| (−5.94) | (−5.74) | ||
| HDI-square | 19.103*** | 19.930*** | |
| (5.98) | (5.78) | ||
| Unemployment rate | −0.325** | −0.405*** | |
| (−2.67) | (−3.71) | ||
| Micro-variables | |||
| Unemployed | −0.074* | −0.013 | |
| (−1.99) | (−1.02) | ||
| Political orientation (right-left scale) | 0.213** | 0.216** | |
| (3.04) | (2.93) | ||
| Education | 0.263** | 0.382*** | |
| (3.06) | (5.96) | ||
| Age | −0.111*** | −0.095** | |
| (−3.80) | (−2.94) | ||
| Inglehart Post-materialism Index | 0.390*** | 0.340*** | |
| (6.57) | (11.87) | ||
| Pseudo-R2 (according to Nagelkerke) | 0.154 | 0.106 | 0.251 |
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| Macro-variables | |||
| HDI | −18.478*** | −19.260*** | |
| (−5.94) | (−5.74) | ||
| HDI-square | 19.103*** | 19.930*** | |
| (5.98) | (5.78) | ||
| Unemployment rate | −0.325** | −0.405*** | |
| (−2.67) | (−3.71) | ||
| Micro-variables | |||
| Unemployed | −0.074* | −0.013 | |
| (−1.99) | (−1.02) | ||
| Political orientation (right-left scale) | 0.213** | 0.216** | |
| (3.04) | (2.93) | ||
| Education | 0.263** | 0.382*** | |
| (3.06) | (5.96) | ||
| Age | −0.111*** | −0.095** | |
| (−3.80) | (−2.94) | ||
| Inglehart Post-materialism Index | 0.390*** | 0.340*** | |
| (6.57) | (11.87) | ||
| Pseudo-R2 (according to Nagelkerke) | 0.154 | 0.106 | 0.251 |
Source: European Values Survey 1999/2000; N=23.826, binary logistic involution, standard error after robust cluster, standardized coefficient, z-values in parentheses, *P<0.05. **P<0.01. ***P < 0.001. The dependent variable has three values as described above: ‘agree with’, ‘do not agree’ and ‘neither-nor’. The choice ‘neither-nor’ was defined in the regression analyses as a missing value.
All of our hypotheses are confirmed, as evidenced by the signs of the standardized coefficients. The more modern the respondent's country, the more likely he is to support EU regulations for equal opportunities between nationals and other European citizens. The respondent's value orientation also affects his conception of non-discrimination: post-materialists support the concept of European-wide equality more than materialists do. The hypothesis that ideological orientation influences attitudes is also confirmed by the empirical analysis. As assumed, the idea of national exclusion is more strongly anchored in right-wing ideology.
Moreover, we assumed that the interests of the interviewee affect his level of support for opening the national job market. Unemployed people as well as people in the lower echelons of the social structure (as measured by the degree of the interviewee's education) are more likely to have to accept disadvantages created by immigration as compared to people who are gainfully employed or have a higher level of education. This assumption bears predictable consequences in regard to citizens’ attitudes toward European-wide equality. In fact, we can show that the greater the unemployment rate in a country, the more citizens support a nationally bounded concept of equality. Consequently, the lower the education of the interviewee and the higher the unemployment rate in a country, the more its citizens tend to discriminate between nationals and other European citizens. The sign of the regression coefficient for individual unemployment points in the right direction, but is not significant. Finally, the control variable age also has the expected effect. The younger the interviewee, the more likely he is to support the idea of equal opportunities on the European job market.13
5 Conclusion
The principle of free movement of persons and labor within the European Union is a central element of the European common market, allowing citizens of any member state to work in any other member state. EU regulation forbids discrimination of European foreigners, stating that domestic citizens as well as citizens from other member states should be given equal treatment. By interpreting the European Law and EU policies, we have first described how the idea of nationally bounded freedom to move, settle, study and work was Europeanized and replaced by an idea in which all European citizens are regarded as equals and therefore can move, settle, study and work in any EU member state.
By analyzing the data from the European Values Survey, we showed that the majority of European citizens still prefer a nationally bounded conception of access to the labor market. The majority of interviewees reject the idea of equal opportunities for nationals and European foreigners on the job market; however, the degree of rejection is quite different in each country. Whereas about 34 per cent of interviewees from old EU member states advocate the idea of non-discrimination, only around 11 per cent of interviewees in the two groups of recent accessions supported this concept. This signifies that EU enlargement substantially decreases the level of endorsement for the EU's non-discrimination policy.
Moreover, clear differences exist at the national level. Support for the notion of non-discrimination is the highest in Sweden, The Netherlands, and Denmark, where well over half of the citizens support European-wide equality. In contrast, less than 5 per cent of Lithuanians, Poles and Maltese support the idea of nondiscrimination. We also analyzed whether the citizens’ attitudes changed over time: in fourteen countries, no change in mentality occurred; support for a nationally bounded market increased in four countries; and support for greater European-wide equality increased in five countries only. Institutional changes as brought about by implementing the freedom of labor have not been paralleled by changes in citizens’ mentalities.
As discussed in Section 2, the indicator we have used to measure the theoretical concept of non-discrimination does not differentiate between immigrants from EU countries and immigrants from outside the EU. Hence, our measurement of the theoretical concept might be biased. Accordingly, the results should be interpreted as preliminary results until better measurements and data sets are available.
The picture drawn in this article looks, however, a little less bleak when one takes into account that citizens from those countries which are most likely to receive migrants are more supportive of the idea of non-discrimination than citizens from countries that are most likely to send migrants. It is possible that people living in countries with a long experience of foreign working immigrants are more tolerant towards foreigners, because they are more familiar with their customs. This might partially explain the differences between EU-15 and Eastern European countries.
The degree to which citizens accept EU regulations is significant in terms of the legitimacy of European policies (Scharpf 2005), in that democracies are structurally dependent on the support of their citizens. A mismatch between an elite project and public opinion can lead to legitimacy problems for EU institutions, as demonstrated by the French and Dutch rejection of the European Constitution referendum.
Following the results of our causal analysis, the modernization process in the new EU countries will play a crucial role for the question of whether citizens’ attitudes will change in the future. We can explain attitudes toward non-discrimination very well with the following theoretically deduced variables: a high level of modernization as measured by the HDI, the interviewee's level of education, and post-materialist values. These three variables have the strongest impact on non-discrimination. One may conclude that support for the EU's freedom of labor and non-discrimination regulations will increase if new member states go through a period of modernization similar to that of old members, and if the modernization time period is not too short. EU membership may accelerate modernization, as was the case for Greece, Portugal, Spain and Ireland (Delhey 2003; Bornschier et al. 2004). These countries were significantly less modernized at the time of their accession, and EU membership has been conducive to modernization.
Footnotes
I would like to thank Joana Schenke for her very in-depth revision of the translation, and especially Silke Hans for assisting with the empirical analysis.
The following data represent the stocks of foreign labour force in different EU countries in 2004 (% of total labour force). They are taken from the ‘International Migration Data 2006’ collected by the OECD. Unfortunately, no differentiation is made between EU and non-EU foreigners. Austria = 11.9%; Belgium = 9.1%; Czech Republic = 2.1%; Denmark = 3.9%; Finland = 1.9%; France = 5.6%; Germany = 9.1%; Hungary = 1.4%; Luxembourg = 62.0%; The Netherlands = 3.8%; Portugal = 5.5%; Slovak Republic = 0.1%; Spain = 6.3%; Sweden = 4.9% and United Kingdom = 5.2%.
Useful information regarding the European Values Survey can be found at http://www.europeanvalues.nl, such as Loek Halman (2001). The data set is available at the Central Archive for Empirical Social Research in Cologne under the number 3811.
Owing to the differing sample sizes in each country, we have weighted the data for the construction of the aggregate categories (EU-15, Enlargement I and II), so that each country gets the same weight.
The variance between the different European countries is rather low: Pearson's correlation between ‘immigrants from richer countries from outside and within Europe’ reaches from 0.90 (Spain) to 0.74 (France); the correlation between ‘immigrants from poorer countries from outside and within Europe’ reaches from 0.95 (Portugal) to 0.78 (Denmark).
The European Social Survey contains only six of the ten 2004 accession countries. Bulgaria, Romania and Turkey are also not included in the ESS. For these reasons, we did not use the ESS as our main data set.
Because the Eurobarometer data set does not include all of the countries under analysis, we forego an extended presentation of the data.
The Human Development Index reaches from 0.742 for Turkey to 0.941 for Sweden.
The Inglehart index was formed on the following basis ‘There is a lot of talk these days about what the aims of this country should be for the next 10 years. Which of the things would you say is most/next most important: (1) Maintaining order in the nation, (2) Giving people more say in government decisions, (3) Fighting rising prices, (4) Protecting freedom of speech’. A new variable was computed forming the Inglehart index: (1) Materialist, (2) Part-Materialist, (3) Part-Post-materialist, (4) Post-materialist.
We used the following question to construct the ‘employed/unemployed’ variable: ‘Are you yourself employed now or not? If yes: About how many hours a week? (If more than one job: only for the main job.) Has paid employment: 30 h a week or more (1), Less than 30 h a week (2), Self employed (3); If no paid employment: Retired/pensioned (4), Housewife not otherwise employed (5), Student (6), Unemployed (7), Other (8)’. A dummy variable was constructed consisting of two values: 0 = employed/pensioner/house wife/student/ other and 1 = unemployed.
The question concerning education was formulated as follows: ‘What is the highest level you have reached in your education? Inadequately completed elementary education (1), Completed (compulsory) elementary education (2), (Compulsory) elementary education and basic vocational qualification (3), Secondary, intermediate vocational qualification (4), Secondary, intermediate general qualification (5), Full secondary, maturity level certificate (6), Higher education – lower level tertiary certificate (7), Higher education – upper-level tertiary certificate (8)’.
The variable was measured in the following manner: ‘In political matters, people talk of “the left”’ and “the right”. How would you place your views on this scale, generally speaking (1 = Left to 10 = Right)’. For the analysis of the data, the variable was reversed: 1 = Right to 10 = Left.
Although the data does have a hierarchical structure (of individuals nested within countries), it is not necessary to calculate a more complicated multilevel model in place of a simpler and more straightforward logistic regression analysis. We took the statistical problems associated with multilevel data into account by estimating robust standard errors. In terms of substance, there is no additional merit to using a hierarchical model because the effects of the micro level variables do not differ substantially between countries – neither in size nor direction. Moreover, there is no theoretical reason to assume that variables like age and education should have different effects in different countries. Therefore, the results of a hierarchical model would be virtually the same as results from a simple regression model, which is much easier for readers to understand and interpret.
In addition we analyzed the European Social Survey data. We calculated a logistic regression model with the question whether the respondent thinks his country should allow people from poorer countries in Europe to live in his country as the dependent variable. The results show, that we can explain the attitude toward non-discrimination between nationals and European foreigners with exactly the same variables we used for our analysis in Table 2. HDI, education and age have a positive impact on citizens’ attitudes towards foreigners, and unemployment rate has a negative effect. Unfortunately the ESS does not contain the Inglehart-Index and the left/right scale to measure the impact of the value orientation and the political orientation of the respondent.
References
Jürgen Gerhards is Professor of Sociology at the Free University in Berlin. His recent books include Shaping Abortion Discourse: Democracy and the Public Sphere in Germany and the United States (together with Myra Marx Ferree, William Gamson and Dieter Rucht), 2002; The Name Game. Cultural Modernization and First Names, 2005; Kulturelle Unterschiede in der Europäischen Union. Ein Vergleich Zwischen Mitgliedsländern, Beitrittskandidaten und der Türkei (together with Michael Hölscher), 2005; Die Herstellung einer öffentlichen Hegemonie. Humangenomforschung in der deutschen und der US – amerikanischen Presse (together with Mike St. Schäfer), 2006; Cultural Overstretch? Differences between Old and New Member States of the EU and Turkey, 2007.