http://eiop.or.at/eiop/texte/1998-009a.htm
:
13.11.1998| This paper's comments page | Send your comment! to this paper |
| Where does supranationalism come from? Ideas floating through the working groups of the Council of the European Union(*) |
| Jan Beyers |
| European Integration online Papers (EIoP) Vol.
2 (1998) N° 9;
http://eiop.or.at/eiop/texte/1998-009a.htm |
| Date of Publication in :
13.11.1998 |
| | Abstract |
Back to homepage |
PDF | PS
|
| This paper's comments page | Send your comment! to this paper | |
1
The central aim of this paper is to explain why some officials involved in the Council working groups are more in favour of European integration than others. There are several reasons for this attention towards the origins of the negotiators' attitudes. First of all, in earlier contributions we showed already that these attitudes are to some degree collectively held. We know that Northern European negotiators were less supranationalist than South Europeans and that the founding nations were more supranationalist than the newer member states (Beyers and Dierickx 1997). These succinct findings and descriptions of attitudes towards integration call however for a deeper and more thorough analysis. Are the factors north/south and founding member-state/newer member-state the only factors that explain the negotiators' position vis-à-vis the integration process? Or do we need to consider other factors as well?
Attitudes towards regional integration concern the level of governance (international, European, national, regional or local) considered to be the most appropriate for the managing policy-problems. Roughly speaking two groups will be distinguished: on the one hand those restraining extensive common policies and solutions and on the other hand those stimulating this by strengthening the role of EU-level policy-making. The first group can be considered as more intergovernmental, the second group as more supranational. Supranationalism is an attitudinal disposition dealing with the question whether or not supranational institutions are primordial to policy-makers or on 'how political authority should be organised within the European Union' (Hooghe 1999). The central dependent variable is thus the locus of management or the place that supranational institutions have in making policies (Sinnott 1995).
There is one important assumption on which this paper rests: that attitudes are important for understanding human behaviour. This is not the same as positing a deterministic relation between attitudes and behaviour. In real life many other factors affect behavioural choices as well, however, a better insight in the origins of ideas, beliefs and attitudes can result in a better conception of policy-making behaviour itself. Moreover, as our data will show, the positioning of nations on the supranationalism-intergovermentalism axis strongly conforms to what well-informed observers of EU-policy-making perceive as trends in the member states' behaviour. There seems to be at least a substantial correlation between attitudes and behaviour(1). Since the central units of analysis are officials and more specifically their individual disposition vis-à-vis the integration process, this paper does not claim to render a better understanding of a concrete policy-events or the policy-making process itself. It is the variation of these individual attitudes on an intergovermentalism-supranationalism scale what we intend to explain.
The intention of this contribution is not the construction of yet another integration theory or a new approach in studying European policy-making. Its objectives are much more limited and modest. In order to understand the reasons why some negotiators are more supranational than others three well-known approaches of studying EU-politics, liberal-intergovernmentalism, political sociology and new-institutionalism will be relied on. From these three schools of thought hypotheses have been chosen and put to an empirical test. Basically three central processes emerge as possible explanations: the principal-agency model viewing officials as member-states' agents, the importance of officials' political values and orientations and finally, the socialisation perspective. Firstly we will delve deeper into these three theoretical tracks. The third section presents the hypothesis to be tested. The fourth section describes the data and the variables to be used and finally, the aforementioned theoretical ideas are confronted with the relevant evidence.
Previous research has shown that nationality is an important factor within the Council working groups (Beyers and Dierickx 1997; Beyers and Dierickx 1998). That officials participating in Council working groups use nationality is not surprising in itself, since these officials prepare the decisions to be taken by their ministers. The Council itself therefore imposes the importance of nationality and formally the officials are delegated negotiators representing mainly national interests.
2
Within liberal intergovernmental thought this conceptualisation has been elaborated extensively (Moravcsik 1991; Moravcsik 1993). It claims that the actors involved in the Council working group work basically under a mandate of their state and that European negotiations evolve in two stages. The national interest is first determined and fixed in a domestic political game and then enters the fray of European negotiations only in a second phase. National representatives are involved in two different games: the domestic game and the European game. Putnam's two-level game has been used extensively to model this complex interaction between European and national decision-making (Putnam 1988).
Consequently the positions vis-à-vis European integration and the attitudes officials hold should be largely a function of some member-state characteristics. Liberal intergovernmentalists, for instance, claim that big countries and countries that hold outlier preferences on public good provisions are relatively reluctant towards further integration (Moravcsik 1993). Smaller countries are considered to be more pro-integration oriented than larger ones, since smaller countries are more intensively affected by the consequences of internationalisation. Within the group of the bigger countries the case of the United Kingdom supports this view, but the case of Italy, however, traditionally a very pro-European member-state, is then conflicting. The Benelux countries have been given as a clear confirmation of the liberal intergovernmental small state hypothesis, though the case of Denmark rejects this view.
Problematic with this line of conceptualising is the fact that it ignores the possible impact of transnational experiences (Risse-Kappen 1997a; Risse-Kappen 1997b), the involvement of international bureaucracies (Atkinson and Coleman 1992) and room for discretion of officials operating within these working groups (Beyers and Dierickx 1997; Egeberg 1996).
Member-state representatives are in fact more than negotiators and diplomats in the traditional sense. The agenda of the working groups deals with complex and very detailed matters and politicians do not necessarily have the time and the ability to instruct officials on every detail. Many of them are technical experts within a specialised field of knowledge (Westlake 1995). Hence, the definition of 'national interest' is not only the result of interactions within member-states, but also, to a considerable extent, the result of interactions between member-states' representatives and supranational actors. Kerremans, for instance, points to the possibility that national interests are eased and that they are reshaped as a consequence of emerging loyalties among negotiators (Kerremans 1996). Such loyalties could be the result of shared political views or convictions. From a similar perspective, Hooghe says that officials operating within the European Commission belong to the more mobilised European citizens and this makes them more sensitive to political discussions in a broader perspective (Hooghe 1997). Therefore one of her research questions is whether Commission officials are affected by cleavages (leftism and supranationalism) in the European public space. Comparatively we can conceptualise the officials involved in Council working groups as more mobilised European citizens and posit a similar research question. Are their attitudes affected by socio-political cleavages in the European public space?
The question is how much room there is for discretion within the working groups? How far are national officials able to pursue their own definition of the situation and thus escape from national or another external role definition? Are working group officials sensitive to ideological partisanship? There are two reasons why we consider a focus on the overall officials' beliefs.
Testing whether officials are affected by cleavages in the European public space starts from the more general idea that EU-politics goes beyond traditional foreign policy among the member states. EU-affairs are consequently structured in a more or less predictable way. This implies that the flow of discrete policy events is part of general political design or project around which broad coalitions emerge (Hooghe and Marks 1997). As such EU-policy-making becomes comparable with policymaking as can be observed within states, which can be understood with distinctions as left-right or centre-periphery.
Secondly, EU-politics is characterised by an oversupply of (often uncertain, imperfect and incomplete) information. The use of ideological concepts functions as a frame of reference for structuring and simplifying this complex world in which officials have to act. Hix states this as follows: 'the shape of the cognitive/ideological environment (the dimensionality of the political space, and the ideological location of the actors within that space) is as much a constraint on political action as 'institutional rules' of the game' (Hix 1998). And Hooghe and Marks claim that ideas are receipts for political action (Hooghe and Marks 1997).
3
The sensitivity to ideological partisanship can be tested by analysing the degree in which supranationalism is related to other political orientations and whether these associations are free from external role definitions. This type of questioning is typical for a sociological way of looking at EU-politics (Hix 1994). Further attention will be on two factors, which are important for scholars working within this tradition: Inglehart's theory of the silent revolution and the debate on the emerging left-right cleavage in EU-politics.
In studying attitudes towards European integration, Ingelhart’s theory of the silent revolution has played an important role. This sociological theory explains attitudes towards European integration by changes in political orientations among the mass public (Inglehart 1977; Inglehart 1990). Changes in the social fabric of societies (e.g. expansion of mass communication and rising levels of welfare) cause value changes at the individual level. Thus instead of attention being given for the so-called traditional materialistic values (e.g. economic growth, economic well being, security and public order) post-materialistic values are becoming more important (e.g. participation in policy-making, human rights, and importance of the environment, quality of life). The theory of the silent revolution links post-materialism with a cosmopolitan attitude and with an evaluation of the nation-state as being too materialistic. Concomitantly post-materialism has, according to Inglehart, a positive effect on support for European integration.
Another political orientation that could affect attitudes towards integration is the left-right cleavage. Since the policy-making of the EU increases conflicts about redistribution and allocation of resources, a socio-economic cleavage will become visible (Hix 1994; Hooghe and Marks 1997). This socio-economic cleavage describes the tension between those claiming that the market is superior to intervention and the regulation of economic affairs. In Western Europe the Internal Market Program and the Economic and Monetary Union provoked and stimulated the debate about how the market should be organised politically. On the one hand, there are those pleading for an integration project which attempts to liberalise the European economies, and refusing to build European institutions capable of regulating the market. On the other hand, there are those demanding for the deepening of the European Union as a means to increase its capacity to regulate capitalism.
Neo-functionalist studies on regional integration emphasised the importance of socialisation within a new emerging supranational polity. Participation in this new polity causes actors to develop new perspectives and definitions of the situation (Lindberg and Scheingold 1970). Several authors argue that the national civil servants involved in the working group meetings are exposed to a spirit of co-operation and mutual understanding, to an esprit de corps (Hayes-Renshaw, Lequesne, and Mayor-Lopez 1989; Hayes-Renshaw and Wallace 1995; Pag 1987; Wessels 1991; Westlake 1995). Thus, member-states representatives are being socialised as European actors and the difference between national and transnational definitions of interests is likely to diminish.
More recently neo-institutionalists have elaborated on socialisation. Their conception, however, differs from the neo-functionalist view in important ways (Bulmer 1994; Bulmer 1997; Kerremans 1996; Pollack 1996). Neo-institutionalists claim that norms and values are embedded within institutions and that individuals learn from their environment what their preferences ought to be (the logic of appropriateness) (March and Olsen 1984). Actors follow 'scripts' and 'templates' and delineate from this the most appropriate behaviour. Such an approach implies that structures affect preferences, strategies and behaviour of social actors. By implication attitudes are dynamic rather than fixed; they change as a result of participation in political decision-making. The questions then are: does participation in the new polity lead to changing perceptions and how do institutions trigger these changes?
Contrary to the neo-functionalist view on socialisation, neo-institutionalists emphasise the incapacity of institutions to shape the values, attitudes and political orientations of social actors. Institutional learning takes time; individuals do not necessarily shift loyalty in response to a functional need(2). Diverse institutional settings compete for the influence over political values and it could be that various institutions affect values and orientations simultaneously. Especially within the European Union with its fragmented multilevel governance the diversity of institutional settings to which political actors belong, makes it rather complicated to disentangle the sources of socialisation (Hooghe 1997; Marks 1996; Marks, Hooghe, and Blank 1996; Risse-Kappen 1996). This neo-institutionalist relaxation of the political spill-over effect is inspired by the recognition that diverse institutions compete for impact on values and political orientations.
4
In her study on the political orientations of Commission officials Hooghe contrasts two sets of factors which are employed in this paper as well (Hooghe 1997)(3). On the one hand, her focus is on experiences within European and other transnational settings. The hypothesis to be tested is whether experiences within these settings stimulate supranationalism. On the other hand, she employs factors to be situated at the level of national political systems (small polities versus big polities, the level of fragmentation of countries of origin and aggregated national elite preferences). The question concerning these factors has a dual nature: firstly, to examine the importance of national level factors vis-à-vis factors describing European and transnational experiences; secondly, to hypothesise on the sign of the association or more concretely; do officials coming from federal systems, small countries or countries with a supranationalist elite support European integration more than others do?
These three theoretical tracks enable us to test several hypotheses, which can be organised under different headings. Table 1 gives a short overview of all the hypotheses to be tested.
First, we can distinguish between theoretical concepts that are operationalised at an aggregate level and which are then disaggregated to the units of analysis' level, in our case, national officials. The officials' nationality and the smallness of the member-state are examples of such aggregate data(4). Second, we present hypotheses concerning individual political values and individual experiences within the officials' national polity. These hypotheses are operationalised at the individual level.
Most hypotheses concern political values or career experiences. Political values and orientations all refer to individual level data. Further on we will make a distinction between transgovernmental experiences and national experiences aiming at testing the relevant importance of both experiences. Some of these hypotheses are operationalised with the help of aggregate data, other employ individual level data.
First, according to the well-known hypothesis of Inglehart, one should expect a positive effect of post-materialism on supranationalism. However, all those scholars that elaborated and commented on Ingelhart's position cannot be ignored. Many came to the conclusion that post-materialistic values are of little value in explaining support for European integration (Anderson and Shawn Reichert 1996; Deflem and Pampel 1996; Dobratz 1993; Duchesne and Frognier 1995; Janssen 1991; Wessels 1995b). Or as some have put it: 'It is difficult to associate post-materialism with the question of potato prices' (Duchesne and Frognier 1995)(5). Therefore we hypothesise no effect of post-materialism with supranationalism.
Second, we expect a positive effect of leftism on supranationalism. Since the European integration project is basically an economic project involving market liberalisation, it could stimulate the support from those aiming at deregulating the economy. In most respects, however, the socio-economic right favours the status-quo when it comes to supranational integration and their support for integration is limited to the reduction of market barriers. Since redistributive policies and market regulation require more governance in general and maybe more European governance in particular it is supposed that leftists are in general more supportive towards supranational integration. This hypothesis conforms the expectation that 'a cleavage ranging from left-leaning supranationalists who support regulated capitalism to rightist nationalists who support neoliberalism' will emerge (Hooghe and Marks 1997).
Finally, we expect that the perception of policy-problems as internationalised has a positive effect on supranationalism (Sinnott 1995). Supranationalism can be considered as a tool for managing the challenges of globalisation. The question is thus whether viewing contemporary policy problems as internationalised increases the chances that respondents consider supra-national institutions as primordial. This hypothesis follows the traditional functional logic of necessity bringing about political integration.
Two sorts of experiences can be distinguished, both focusing on experiences in transnational settings: transgovernmental experiences and the more specific experience negotiators had with EU-affairs. These experiences are supposed to have a positive effect on supranationalism(6).
5
There are different sorts of transgovernmental experiences, which could have a different impact on supranationalism. On the one hand, an official could have career experiences within an international organisation or a European institution(7). We can imagine that such experiences result in a positive disposition towards supranationalism. On the other hand, an official could have worked within a diplomatic representation or an embassy. Consequently we can question whether such experience is really transgovernmental and whether it will strengthen supranationalism. One could equally presume that these experiences strengthen intergovernmentalism since diplomats are officially assumed to be experts in representing and defending national interests abroad.
Furthermore one could argue that transgovernmental socialisation as such is not that important because the transgovernmental setting of the European Union is so manifestly different from other transgovernmental settings that the unique experience of being involved in European affairs counts more than any other factor. By the time one gets more involved in the peculiarities of European policy-making, one learns that political processes within the European Union are substantially different from proceedings going on in traditional international organisations or from the policy-making processes in domestic settings. The consequence is that the traditional implementing role of diplomats and civil servants will transform gradually into a policy-making role and that the definition of 'national interest' will become more dependent on interactions taking place within the working groups than being uniquely dependent upon domestic interactions.
A reading of recent and less recent articles on European policy-making points to the fact that many authors employ the North-South distinction as a tool for understanding EU-politics(8). One question closely related to the North-South issue concerns whether perceived economic benefits stimulate supranationalism. In essence the European integration project is still heavily an economic project for which support can be evaluated in the light of the profits participants get from it. The Internal Market Program and the EMU originated from a bargain between countries benefiting from a large free trade area and those benefiting from the extension of redistributive structural policies(9). The fact that the importance of supranational integration in the sense of an increased redistribution between 'have' and 'have-nots' is in the advantage of the Southern states should stimulate an overall positive disposition towards supranationalism.
Following Hooghe we hypothesise that negotiators coming from small polities may favour supranationalism more than negotiators representing the bigger member-states (Hooghe 1999). Smaller states' representatives know that their country is more intensely affected by the consequences of internationalisation and therefore they adapt themselves more to the fact that national policy alternatives are largely constrained. Consequently supranationalism is an option for these countries.
Another hypothesis to be found in the literature is that the public of founding member states has a more supranational attitude than the newer member-states (Anderson and Shawn Reichert 1996; Eichenberg and Dalton 1993). Similarly one could argue that the longer a state is member of the European Union, the more experiences its civil service has had with it and that this makes the internalisation of supranational values and norms more feasible.
The weakness of such a simple dichotomous concept is that is makes abstraction of the time spirit and the motivation of the candidate at the moment of entrance (Anderson and Kaltenthaler 1996; Hooghe 1999). The founding member-states (the Benelux, France, Germany and Italy) originated the European institutions during a period characterised by a permissive consensus and they had more socialisation opportunities, that is years to learn the peculiarities of European policy-making. States that entered in the first wave (United Kingdom, Ireland and Denmark) were at the moment of their entrance reluctant toward supranationalism and we assume, therefore, that these member-states did not appoint supranational officials in the working groups. On the contrary, there are good reasons for assuming that they charged the most intergovernmental ones with representation and negotiation tasks. Although these officials had more time to become socialised than those whom entered the EU at a later stage; they started off as more intergovernmental than the Southern states (Spain, Portugal and Greece) which became members in the second wave. The latter favoured supranational policies (e.g. the extension of social and structural policies) from the very beginning.
Hooghe hypothesises that the views of Commission officials may be a reflection of values shared by the respective national elites (Hooghe 1999). Indeed most officials received their training and political education within national political systems. Thus, national civil servants and diplomats learned the appropriate values and orientations within the national elite to which they belong.
6
Finally, the structure of the national polity and whether constitutional features, territorial decentralisation and the role of regions in the central government affect the officials' view will be analysed. The hypothesis is that negotiators representing federal polities have much less difficulties in adapting themselves to a supra-national way of policy-making, since they are used to operating within a multilevel governance system at home. Hooghe says: 'EU governance merely adds another protective layer of structuring, which pushes back the uncertain external environment.' (Hooghe 1999). As a consequence socialisation takes less time and effort.
There is another reason why federalism could stimulate supranational attitudes. One could assume that negotiators representing federal systems are more dependent on complex domestic co-ordination procedures than those representing more unitary systems. The former have the difficult task to take multiple interests into account. In such a situation supranationalism eases the negotiators' work since it expresses a good deal of diffuse support for the initiatives and proposals of the European Commission. As a result, less co-ordination with, and control by the various domestic authorities is needed (Dierickx and Beyers 1999). After all, a defensive stance against the Commission, such as the British one, requires more effort, more co-ordination and more control, a situation that officials from federal polities will try to avoid.
It is important to note that most hypotheses are framed, with exception to the North-South hypothesis, within a socialisation logic. Admittedly some of these concepts might also fit within the logic of viewing officials as agents of national interests. This is especially the case for size and national elite orientations.
A positive effect of smallness on supranationalism means that smallness constrains the range of national options and leads to a limited esteem for sovereignty. Socialisation implies then that officials coming from smaller states have learned that supranationalism is an adequate adaptation to externalities. A positive effect of smallness is also expected within liberal-intergovernmentalism, which focuses on the fact that 'national interests would lead one to expect large, self-sufficient and uncompetitive countries… to be relatively unwilling to accept stronger supranational institutions…' while 'smaller countries might be expected to support strong supranational power.' (Moravcsik 1993). To conclude, despite its interpretation (socialisation of defending the national position) the expected effect of size has to be positive for both approaches.
A similar situation exists when it comes to the overall national elite orientation. A national negotiator can also be considered as an agent to whom national politicians have delegated the competence to perform tasks in accordance with orientations around which a national political elite has found a national elite consensus(10). Again, whether we interpret this effect as a socialisation effect or from purposive negotiators' instructing, the sign of the effect should be positive, that is the more a national elite favours supranationalism the more negotiators with the same nationality are pro supra-national integration.
Only the North-South hypothesis is framed within a conception of negotiators representing hard-edged national economic interests. Nevertheless, it is possible to understand the North-South hypothesis from a socialisation logic as well. Egeberg, following Hofstede, hypothesises that Northern Europeans are culturally more adapted to Weberian forms of administration, which implies that they are more capable to make abstraction of idiosyncratic socio-cultural characteristics (Egeberg 1996; Hofstede 1991; Hofstede 1994). Southerners are more attached to group norms and define themselves more as members of collectivities, which could imply that they assign more weight to preferences of their mother country and to solutions that are successful at home. This logic could lead to a less supranational position and thus an effect that runs counter to the theoretical expectations we proposed.
The structure of the national polity and the length of membership are easy to interpret from a socialisation process. Therefore it is important to test whether these two concepts have an effect on supranationalism and whether this effect is really distinguishable from the other more ambiguous concepts. If this is the case, we have a more straightforward indication of the fact that socialisation is at work.
7
In contrast with the hypothesised positive effect of transgovernmental experiences on supranationalism, we suppose that the longer officials served in a national administration the lower their supranational disposition. Service in a national administration increases the chance that a negotiator internalised the values inherent to a domestic administrative and political system (Hooghe 1998).
The final hypothesis to be tested is the innovator/escapists hypothesis (Martinotti and Stefanizzi 1995). The idea is that some negotiators are critical about the deficiencies of their own political system and that they therefore develop a positive expectation vis-à-vis the European integration. This idea has been explored in public opinion research and it appeared in other studies as well (Anderson 1995; Hoffmann 1966; Hooghe 1999; Martinotti and Stefanizzi 1995). A traditional example in this respect is Italy, were both the elite and public opinion is largely supportive towards European integration and where the state is relatively weak and considered to be unstable(11).
In the first half of 1994 Belgian civil servants representing their country in, at that time, about 170 working groups of the Council of Ministers, were interviewed(12). About 110 Belgian civil servants from specialised ministries participated in the working groups on a part-time basis, which implies that they only occasionally dealt with the Council working groups. To this should be added a total of 21 diplomats and 13 civil servants who were part of the Belgian Permanent Representation, and also some 15 diplomats from the Ministry of Foreign Affairs, in all 49. At the time of the interviews the 13 civil servants belonging to the Belgian Permanent Representation functioned as liaison officials for the federal ministries of Agriculture, Finance, Transport, Development Aid, Foreign Trade, Economic Affairs, and Environment. These 49 diplomats and civil servants of the Permanent Representation and Foreign Affairs are involved in Council working groups on a full-time basis. With respect to European policy-making the full-timers have been more exposed to the peculiarities of the European policy-making setting than part-timers. Of this estimated population of 159 Belgian civil servants and diplomats (110+49) a sample of 65 part-timers and 30 full-timers was drawn.
The total number of diplomats and civil servants in the Permanent Representations of the (then) 12 member states can be estimated as 300. The third sub-sample consists of 108 respondents representing the other member states of the European Union in the same 13 working groups as the Belgian full-timers.
All respondents mentioned in Table 1 were interviewed with the help of a standardised questionnaire(13). The sampling design allows three important comparisons to be made. Firstly, full-timers can be compared with part-timers. Secondly, the Belgians are compared with their foreign colleagues who participated in the Council working groups. Thirdly, within the multinational part of the sample a comparison between North and South Europeans, between small and big member states and between older member states and newer member states can be made.
The subsequent multivariate analyses will be carried out on two sub-samples: on the one hand we consider a multinational group sampled on basis of their involvement in the 13 selected working groups and on the other hand the Belgians are considered as a separate sample. The first sub-sample matches the first column and the second sub-sample is composed of the first row of Table 1.
Three items (see Table 3) were employed in order to measure supranationalism(14). They deal mainly with the proceedings within the working groups of the Council of Ministers and imply the strengthening of European policies and the European institutions in general. An objection could be that these items do not tap the functioning of concrete institutions (for instance the Commission or the Parliament). However, we assume that since these items concern the concrete task and mission a Council official has to perform, they fit well within the concept that we intend to measure.
8
The stimuli all appeal to the willingness and preparedness to strengthen EU-level policy-making and the relative autonomy of the EU-level vis-à-vis the member-states. This measurement instrument does not assume à priori that the Council is by definition something intergovernmental. Indeed some elements in the Council proceedings are more intergovernmental (e.g. when unanimity voting is required) than others (e.g. when qualified majority voting is possible). Whether the Council functions in a more intergovernmental or a more supranational way is, according to our viewpoint, also dependent on the national representatives' attitudes. Some are in favour of more extensive common policies, while others restrain this.
A first look at the univariate distributions in table 3 shows a rather high degree of dissensus among the non-Belgians concerning proceedings within the working groups. Roughly speaking two groups can be distinguished: on the one hand those restraining extensive common policies and solutions and those stimulating this by strengthening the role of EU-level policy-making. The Belgians clearly belong to the second group. A principal factor analysis was carried out in order to find out whether the three items belong to one dimension(15). In order to retain robust solutions the analysis was carried out for the two sub-samples separately. Table 4 shows the loading patterns of a principal factor analysis resulting into one factor.
Postmaterialism: The first two items in
Table 5 semantically fit with what
Flanagan called the authoritarian-libertarian cleavage (Flanagan 1987) and
the second two items are more linked with Ingelhart's conception of
postmaterialism (Inglehart 1977). Belgian full-timers are the most
materialists (
=-0.31)
and the differences between the non-Belgians and the Belgian part-timers
is negligible (averages are respectively 0.03 and 0.09) (16).
Table 5 Table 6 Table 7 Table 8
Leftism: For leftism three items are presented in table 6. Dierickx and Majersdorf used these items before (Dierickx and Majersdorf 1994) and we employed them in another paper as well (Beyers and Dierickx 1997). With respect to leftism we did not find significant differences between Belgians and non-Belgians or between part-timers and full-timers.
Policy-perception: The five items presented in
Table 7 concern the way respondents
perceive policy-problems as Europeanised(17).
Competencies of the European Union are recognised insofar as the European
character of policy-problems is identified. On average the Belgian sample
is more convinced that policy problems are internationalised than the
multi-national sample (=-0.27
for the part-timers and =-0.35
for the full-timers). For the non-Belgians the average is +0.26.
Transgovernmental experiences: The number of years a respondent was involved in the various working groups is taken as an indicator of transgovernmental experiences. For the non-Belgians this is on average 9 years, for the Belgian full-timers 12 year and the Belgian part-timers 11 years.
Size: Size is measured by a simple dichotomous variable dividing the multinational sample into small and big countries. Respondents representing a country with a population smaller or equal to 15 million were labelled small (Ireland, Denmark, Belgium, Netherlands, Luxembourg, Portugal and Greece, N=63), others were labelled big (France, Germany, United Kingdom, Spain, Italy, N=57).
North-South: For North-South we have chosen the distinction of Hofstede (Hofstede 1991; Hofstede 1994), which divides the multinational sample into ‘North Europeans’ (Germans, Danish, Dutch, British, Irish and Luxemburgers, N=55) and ‘South Europeans’ (Belgian, French, Portuguese, Spanish, Italians and Greek, N=65).
Length of membership: We construct a theoretical variable which posits a priori that supranationalism should decrease in the following way. The founding member-states should express themselves as the most supranationalists, the states that entered in the second wave are less supranationalist than the founding members, but considerable more supranationalist than the states that entered in the first wave. Concomitantly, the original members received a code 1, those belonging to the second enlargement were coded 2 and the first wave received a 3, indicating the rising level of intergovernmentalism to be expected.
9
National elite orientation: For testing whether the views expressed by our respondents are similar to those shared by the national elite we propose, in order to make results comparable, exactly the same indicator as Hooghe, which is based on the Eurobarometer elite survey of 1996 (see Table 10) (Hooghe 1999).
Federalism: The last aggregate variable is again similar to the variable Hooghe employed for describing the extent to which a polity is federalised (Hooghe 1999). With this index (see Table 11b) countries are coded on basis of constitutional characteristics, territorial decentralisation and the role of regions in the central government. A high score (7) indicates an extensive level of federalism, while a low score (0) describes more centralised and unitary states.
Organisational self-esteem: The disposition vis-à-vis the national polity is made operational with the help of the scale organisational self-esteem. The relevant items and factor-loadings are presented in Table 8. Elsewhere the semantics of this scale and its relation with administrative culture were discussed more fully (Beyers and Dierickx 1997; Dierickx and Beyers 1999). From previous research we know that the organisational self-esteem is very low among the Belgians and especially among the Belgian part-timers.
Work experiences in the national polity: Finally we take the number of year a respondent served the national administration as a measure of domestic career experiences. Non-Belgian negotiators served their state on average 16 years, while for the Belgian full-timers it is 21 years and the Belgian part-timers it is 19 years.
The proposed hypotheses are examined in two steps. First, we look at the bivariate relations between the independent variables and supranationalism. This data-analysis has some disadvantages and therefore we propose to employ in a second stage an analysis of co-variance (ANCOVA), which takes interaction-effects more explicitly into account.
We begin with political values and orientations. First, as recent public opinion research does, Ingelhart's hypothesis must be refuted. Furthermore, the bivariate analyses looks promising for those hoping that the strong European involvement of national officials will evolve towards a European public space. Although the correlations are not that strong, it seems that within the multinational sample leftism and an evaluation of policy problems as internationalised are positively related with supranationalism. In the Belgian sample the perception of policy problems as internationalised has a positive association with supranationalism.
It turns out that transgovernmental experiences are not a factor to be considered. In both sub-samples the relation is weak and not significant. This points to the fact that negotiators' attitudes are not necessarily shaped by long experiences in EU-affairs.
National factors proved to be the most predictive. Southern Europeans
are more favourable towards supranational solutions (=-0.09)
than their Northern European colleagues (=0.68).
And as expected the founding member-states' representatives (=-0.12)
are still more attached to the ideal of European integration than the
newcomers (=0.60). The
hypothesis that the period of entrance has an impact is confirmed by a
strong correlation. Representatives of states that entered in the second
wave are considerably more supranational (=0.18)
than those that entered in the first wave (=1.02).
Also the relation between the national elite orientations and the respondents' attitudes is strong. Table 10 gives an overview for all nationalities in the multinational sample and relates the data to the period of entrance. In general most nationalities fit within the general trend to be discovered by our bivariate data-analysis. The Belgians and Italians belong to the most convinced supporters of supranationalism and the British and the Danish are strongly attached to intergovernmentalism. There are two deviations for which we do not have a clear explanation. The less pro-European stance of the Portuguese national elite is not reflected in our sample where the Portuguese belong, as expected by some of our hypotheses, to the supranational camp. The opposite holds for the Germans. The German national elite belongs, as supposed by several of our hypotheses, to the supra-national group, but this is not confirmed for the German officials(18).
The bivariate relations with organisational self-esteem and experiences in the national civil service confirm that national socialisation experiences are important. But the sign of the relations is not always what we expected. The analysis shows, as hypothesised, that a low organisational self-esteem stimulates supranationalism. This is especially the case for the Belgian sample. However, the longer an official has worked within national settings, the more he favours supranationalism, a result which runs counter our expectations.
10
Finally, before we turn to a co-variance analysis we would like to focus the readers' attention on two variables, federalism and size, that had a significant and substantial effect in Hooghe's research on Commission officials (Hooghe 1999). In our sample, these variables are insignificantly associated with supranationalism.
Respondents negotiating on behalf of small states express themselves
only slightly more pro-European (=0.10)
than those representing bigger states (=0.44).
Table 11a and
11b shows the levels of
supranationalism for diverse levels of federalism. More supranationalism
should be present as the level of territorial decentralisation increases.
Our data do not confirm this idea though. It is clear that the three most
federal polities (Germany, Spain and Belgium) express themselves the most
supranational, but there are considerable differences among these three.
Furthermore, respondents of some non-federal polities (France and Italy)
situate themselves more on the supranational than on the intergovernmental
side and unitary systems are barely more intergovernmental than
non-unitary polities.
Before we turn to the results of a co-variance analysis we will show with some simple examples the misleading character of the previous analysis.
First of all, interaction effects. Remember that we concluded that territorial decentralisation or federalism and size are not related to supranationalism. It could be, however, that the overall relation between territorial decentralisation and supranationalism fades out as a consequence of the fact that the strength of this relationship is affected by the size of the respondents' polity. The smaller a country, the more its decentralised character will affect supranationalism and the bigger a country the lesser decentralisation will be related to supranationalism. This means, we hypothesise, an interaction-effect of size and federalism on supranationalism. A first glance at the correlations between federalism and supranationalism within the subgroups of bigger and smaller countries confirms the idea that federalism is positively related to supranationalism (r=0.36, p=0.0041) within the former subgroup and that is has no importance in the latter (r=0.08, p=0.5443)(19).
A similar phenomenon is observed for policy-perception and size. Within the subgroup of respondents coming from the smaller countries we observe a significant and very substantial relation between policy-perception and supranationalism (r=0.50, p=0.0001), while for the bigger countries this relation is completely absent (r=0.02, p=0.8715). Thus, respondents of the smaller polities are more convinced that policy-problems are internationalised and they associate this with supra-national decision-making styles in the council. For the bigger member-states this seems not to be the case.
A traditional OLS-regression model ignores such substantively interesting interaction effects. An advantage of co-variance analysis is that it enables us to test these interaction-effect.
Second, disaggregating aggregated data has the disadvantage that it can lead to misinterpretations of results. In Table 12 we compare the correlations of the 5 macro-variables for aggregated and desaggregated data. Part A shows the association of a measurement at the respondents' level (N=120) and part B the correlations for the aggregated data (N=12, which is the real measurement level). In general, correlations in the multinational sample are higher than in the aggregated sample and even more important are the decreasing p-values. The lower p-values in the multinational sample are a consequence of the lower standard errors, which are in turn a result of the fact that the number of observations is artificially increased by disaggregating data. We tend to accept, therefore, relations between variables at the respondents' level too easily, relations which do not exist in a similar magnitude at the real measurement level of these variables. To put it differently, there is a danger of overestimating the effect of especially nation-specific variables while underestimating individual level variables.
Finally, aggregate data leads to the danger of measuring very similar empirical aspects with different operationalisations. In table 12 the national elite-orientations are strongly related to the period of entrance. In the founding member states the national elite is still more supranationalist than in the newer members states. Among the latter a distinction can be made between the first wave and the second wave of which the latter are more supranationalist than the former. This is the case for both aggregate and the desaggregated data and it implies that we have to be careful with an analysis carried out with these two independent variables since they risk to create problems of collinearity. The same observation can be made for the North-South distinction, which is strongly associated with the period of entrance and national-elite orientations. The second enlargement was only composed of Southern European states and the first consisted only of northern countries. These features of the data force us a) to limit our ambitions for the multinational sample to carefully chosen independent categorical variables, namely size, period of entrance and federalism, and b) to be cautious with the interpretation of effects.
11
It is important to keep in mind that the variance on supranationalism is substantially lower within the Belgian sample than in the multinational sample. The standard deviation equals 1.07 in the multinational sample, while the same coefficient is only 0.79 in the Belgian sample. Because of the relative pro-European and homogenous character of the Belgian sample the proportion explained variance is not expected to be very high.
For the Belgian sample we also make a distinction between full-timers and part-timers, two groups that are substantially different with respect to their working group experiences. Consistent with the hypothesis on transgovernmental experiences we expect that full-timers are more pro-European than part-timers.
Three models were tested (see Table 13). The significance of model 1 and 3 does not differ very much, so we will interpret the first-order effect of policy-perceptions and the interaction effect of involvement and organisational self-esteem. Table 14, part A displays the averages of supranationalism by policy-perceptions and part B shows the average supranationalism by working group involvement and organisational self-esteem.
This analysis confirms largely the bivariate analysis, which showed that two factors have to be taken into account when we want to interpret Belgian supranationalism. The first factor points to the role of policy-perceptions. This observation runs parallel to an observation we will make for the multinational sample, namely that representatives of small polities do in general perceive policy problems as more internationalised and in doing so they will be more supranational. The Belgians are no exception to that. The more they conceptualised policy problems as internationalised, the more they expressed themselves in favour of a supranationalist way of policy-making.
The second factor points to the importance of a national socialisation factor, namely organisational self-esteem. This factor has a bivariate (see Table 9) and a first order effect on supranationalism (see model 1 Table 13). In general a low organisational self-esteem stimulates supranationalism. For a profound substantive interpretation, however, we take the amount of experiences with EU-affairs into account. The involvement within the working groups as such does not explain supranationalism, but the relation between organisational self-esteem and supranationalism changes as a consequence of involvement in the working groups. As can be seen in part B of Table 14 the lesser an official was involved in the working groups, the more his low self-esteem goes together with supranationalism.
For the multinational sample two models were tested. In model 1, without interaction terms, only two factors had an effect: work experience in national settings and period of entrance. According to the second model we have to interpret three interaction terms of which one is a third-order-effect. For the interpretation of the results we turn to Table 16, part A to D.
Firstly, the interaction effect of policy-perception and size. Officials coming from the bigger countries do not become more supranational when their perception of policy problems as internationalised grows. In the case when a respondent represented a smaller member state, the characterisation of policy problems as internationalised stimulated supranationalism substantially. Thus the smaller member-state attitude can be explained in a similar vein as the Belgian attitude.
Secondly, the interaction of size and federalism on supranationalism is a bit more complicated because the interaction-effect does not follow a clear linear trend. On the one hand, the data show that federalism as such does not necessarily fit with supranationalism. Representatives of smaller polities are more supranational if they belong to a federal system (the Belgians) but this result is mixed for the unitary systems: Dutch, Luxemburgers, Greek and Irish are more supranational but Danish respondents are, despite their smallness, more intergovernmental. The bigger non-unitary member-states show a more supra-national disposition than British respondents. So here it seems that territorial decentralisation strengthens supranationalism. However, Germans and Spanish express themselves as less supranationalist than we expected on theoretical grounds.
12
Finally, we have a rather complex third-order-effect. In order to hold the interpretation accessible we concentrate on the second-order-effects, which are encapsulated in the third-order effect. The third-order-effect simply points to the fact that the two effects do not contradict each other.
The hypothesis of the founding member-states as the most supranationalist, the states that entered the EU in the second wave as holding an intermediate position and the first wave as the most intergovernmentalist is confirmed. However, within the three waves there are considerable differences when it comes to national career experiences. A long experience as a national civil servant stimulates supranationalism for the founding member states. For respondents belonging to the second wave the difference is less outspoken. The opposite is observed for negotiators belonging to the first wave: less experiences in domestic administrations attenuates intergovernmentalism. We obtain a similar picture when we look at the working group experiences. For those belonging to the first wave, working group involvement does not relax their intergovernmentalism, while it stimulates the supranationalism of the founding member states and the second wave.
This leads to two conclusions. Firstly, our hypothesis concerning the period of entrance is confirmed. Founding member-states have had more socialisation opportunities in European affairs and their representatives are consequently convinced of supranationalism. For representatives of the first wave, a limited exposure to the national administration weakens intergovernmentalism, but extensive involvement in the working groups does not stimulate more supranationalism. This confirms the hypothesis that the transgovernmental socialisation opportunities of those belonging to the first wave are constrained. In comparison to the first wave, socialisation turns out to be more successful for the second wave.
Secondly, we need to reformulate the initial hypothesis concerning the negative impact of domestic career socialisation on supranationalism. To start with, involvement in the working groups is related to national career socialisation. This points to the fact that transgovernmental socialisation opportunities in EU-affairs are likely to depend on career socialisation within national polities(20). Only for negotiators of the first wave experiences in national settings attenuate intergovernmentalism. National socialisation matters, but it does not necessarily matter in the sense as hypothesised.
Furthermore, the strength of involvement in the working groups should not be overestimated. This can not be directly inferred from our analysis with the multinational sample. Within the Belgian sample we observe the importance of the national socialisation factor in the form of organisational self-esteem. This factor stimulates supranationalism among those that are less exposed to EU-affairs (the part-time officials). Thus a negative disposition towards the national apparatus has a positive effect on supranationalism, and this effect is more outspoken among those with lesser European experiences.
Existing political orientations do not substantially influence supranationalism. This is reflected in the negligible effect of leftism and post-materialism. Only the perception of policy-problems as internationalised seems to play a role. However, this effect is only substantial in the case of the smaller member-states and is irrelevant for bigger member-states. Therefore this interaction effect does not fit within a sociological way of understanding the officials' attitudes. It remains unclear whether this effect has to be considered as a result of an institutional socialisation logic or an intergovernmental logic stating that officials are agents of national interest.
Why do political orientations have no effect on supranationalism? One problem with our operationalisation is that it probably neglects the unique and complex character of European integration. The theory of the silent revolution ignores the fact that the European integration project is largely an economic project; the role of citizens and democracy should not be overestimated. Our research also confirms the hypothesis that leftism is of little use in describing attitudes towards European integration (Hewstone 1986; Taggart 1998; Wessels 1995a). The only political orientation that matters is whether political actors define policy problems as internationalised or not, but this effect is largely dependent upon the state one represents.
Supranational attitudes are substantially affected by an interaction between domestic and transgovernmental experiences. The analysis shows that an extensive career within the national administration does not necessarily have a negative effect on supranationalism. On the contrary, we found that in some occasions the officials with more domestic administrative experiences proved to be more supranational than those with less national experience. This effect interacts with the period a member state entered the European Union and the socialisation opportunities this created for negotiators. For the founding member-states a long experience in the national civil service strengthens the transgovernmental socialisation opportunities in EU-affairs, which in turn has a positive effect on supranationalism. The respondents belonging to the first wave (the Irish, British and Danish) are still affected by a negative mood towards integration and consequently intense involvement in working groups does not have the expected positive effect. On the contrary, in this case those with limited national experiences are a bit less intergovernmentalist. Finally, negotiators belonging to the second wave are positively affected by transgovernmental socialisation opportunities. In their case experiences in national settings do not matter very much and the number of years involved in working groups stimulates supranationalism.
13
Furthermore, the structure and the size of national polities matters for supranationalism. It is however more size, in interaction with policy-perception, which convinced us of this than federalism. We cannot conclude that those used to less centralised decision-making within their home countries are more attracted by supranational decision making. In order to understand supranationalism fully we have to conclude that the national political and administrative culture has to be taken into account. This is reflected convincingly in the Belgian sample where a negative disposition vis-à-vis the national polity results in more supranational positions. This observation points to a socialisation logic that is at work and it fits within the neo-institutional view that socialisation is not necessary a functional process. It is not only because Belgians consider the EU as more efficient or effective that they are supranational. It is also a matter of their socialisation within the national political and administrative apparatus.
The latter observation, however, does not imply that the formal institutional structure of the national polity does not matter. We infer this from considerations about the national polity made extensively by Belgian officials in 23 qualitative interviews. These Belgians emphasised that the constitutional reforms in Belgium had the most pervasive impact on their work. In contrast, the increasing impact of the European institutions was mentioned in only one interview. These qualitative interviews also showed that constitutional reforms are experienced negatively. Especially in the Belgian federal administration complaints about the increased complexity of the administrative and political system prevailed: the interplay between different agencies is not efficient or absent; the involvement of a large number of ministries ‘increases the probabilities of bad instructions’. Another indication of the impact of the structure of national polities on views about EU-governance comes from our Italian respondents. Italians belong to a big, non-federal state, and they should therefore dispose supranationalist. Together with the Belgians, however, the Italians belong to the most supranational group and although their constitutional systems are very different, both belong to very fragmented polities.
The non-interpretable effect of federalism implies that the hypothesis concerning federalism needs careful reconsideration. Are negotiators belonging to federal polities indeed convinced that multilevel governance is more effective and are they really more accustomed to policymaking at multiple levels? The scale of organisational self-esteem is a good starting point for elaborating on this. Belgians belong to one of the most decentralised and federalised polities of Western Europe. However, they are not unanimously convinced that their national multilevel system is efficient and effective and we cannot conclude that their political culture is well adapted to policymaking at multiple levels (Dierickx and Beyers 1999). Equally we can hypothesise that Italian respondents, for instance, are more supranational, not in spite of the absence of federalism, but because they consider their system as too centralised and thus, according to their views, as lacking efficiency and effectiveness.
In fact the problem with measuring federalism is also linked with the complications of using aggregate data. Is the federalism index an adequate operationalisation of polity-fragmentation? And more importantly, does it say something about how negotiators perceive their domestic polity? For fragmented (or small) polities supranationalism could be an option in order to structure an uncertain environment (Hooghe 1999), but it remains unclear whether individual negotiators are also convinced about this. Are officials from federal polities really used to rule-making at multiple levels? Hence, federalism is not necessarily a sufficient indicator of fragmentation. In fact when we employed federalism as an indicator we consider this contextual feature as an individual attribute of a respondent. The question however must go deeper and concentrate on the intervening mechanism between social structure and individual attitudes.
Finally, we must admit that our inquiry has still an explorative character and that probably more data is needed in order to confirm or falsify the proposed hypotheses. It would have been formidable if we disposed a similar sample of about 60 part-timers for all member-states. Comparisons and measurement would have been more elaborate and from a data-analytic viewpoint more sophisticated testing would become possible. Of course every research undertaking is partly an individual effort, but we think that more transnational links and talks could stimulate progress. For the sake of comparability we borrowed extensively from Hooghe's very interesting paper, but while doing this we regretted that we did not employ the same items and measurements in order to make results even more comparable. We did not include the variables that Hooghe used for testing principal-agency hypotheses, although this factor could have been more important for our sample (member-state officials) than for her sample (Commission officials). In a similar manner we employed factors as organisational self-esteem which could have been promising for Hooghe's sample. Let these final words not only be a regret or a critique, but rather a plea for more cumulative empirical research by strengthening transnational co-ordination and co-operation.
Anderson, Christopher. 1995. “Economic Uncertainty and European Solidarity Revisited : Trends in Public Support for European Integration.” Pp. 111-133 in The State of The European Union. Building a European Polity, edited by C. Rhodes and S. Mazey. Harlow: Longman.
Anderson, Christopher J. and Karl C. Kaltenthaler. 1996. “The Dynamics of Public Opinion toward European Integration, 1973-1993.” European Journal of International Relations 2:175-199.
Anderson, Christopher J. and M. Shawn Reichert. 1996. “Economic Benefits and Support for Membership of the EU : A Cross-National Analysis.” Journal of Public Policy 15:231-249.
Atkinson, Michael M. and William D. Coleman. 1992. “Policy Networks, Policy Communities and the Problems of Governance.” Governance 5:154 - 180.
Beyers, Jan and Guido Dierickx. 1997. “Nationality and European Negotiations : The Working Groups of the Council of Ministers.” European Journal of International Relations 3:436-471.
Beyers, Jan and Guido Dierickx. 1998. “The Working Groups of the Council of Ministers : Supranational or Intergovernmental Negotiations?” Journal of Common Market Studies 36:289-317.
Bulmer, Simon J. 1994. “The Governance of the European Union : A New Institutionalist Approach.” International Public Policy 13:351-380.
Bulmer, Simon J. 1997. “New Institutionalism, The Single Market and EU Governance.” ARENA working paper 97/25, Oslo.<http://www.sv.uio.no/arena/publications/wp97_25.htm>
Christoph, J.B. 1993. “The Effects of Britons in Brussels : The European Community and the Culture of Whitehall.” Governance : An International Journal of Policy and Administration 6:518-537.
Deflem, Mathieu and Fred C. Pampel. 1996. “The Myth of Postnational Identity : Popular Support for European Unification.” Social Forces 75:119-143.
Dierickx, Guido and Jan Beyers. 1999. “Belgian Civil Servants in the European Union. A Tale of Two Cultures?” West European Politics :forthcoming.
Dierickx, Guido and Philippe Majersdorf. 1994. De Politieke Cultuur van de Ambtenaren en de Politici in Belgie. Brugge: Vanden Broele.
Dobratz, Betty A. 1993. “Changing value orientations and attitudes toward the European Community : a comparison of Greeks with citizens of other European Community nations.” East European Quaterly 27:97-127.
Duchesne, Sophie and André-Paul Frognier. 1995. “Is There a European Identity ?” Pp. 193-226 in Public Opinion and Internationalized Governance, vol. 2, Beliefs in Government, edited by O. Niedermayer and R. Sinnott. New York: Oxford University Press.
Egeberg, Morten. 1996. “Organization and Nationality in the European Commission Services.” Public Administration 74:721-735.
Eichenberg, Richard C. and Russel J. Dalton. 1993. “Europeans and the European Community : the dynamics of public support for European integration.” International Organization 47:507-534.
Flanagan, S.C. 1987. “Value Change in Industrial Societies.” American Political Science Review 81.
Gabel, Matthew and Harvey D. Palmer. 1995. “Understanding variation in public support for European integration.” European Journal of Political Research 27:3-19.
Garrett, G. 1992. “International cooperation and institutional choice : the European Community's internal market.” International Organization 46:533-560.
Hayes-Renshaw, F., C. Lequesne, and P. Mayor-Lopez. 1989. “The Permanent Representations of the Member States to the European Communities.” Journal of Common Market Studies 38:119 - 137.
Hayes-Renshaw, Fiona and Helen Wallace. 1995. “Executive power in the European Union : the function and limits of the Council of Ministers.” Journal of European Public Policy 2:559-582.
Hewstone, Miles. 1986. Understanding attitudes to the European Community. A social-psychological study in four member states, Edited by R. J. Eiser and K. R. Scherer. London: Cambridge University Press.
Hix, Simon. 1994. “The study of the European Community : The Challenge to Comparative Politics.” West European Politics 17:1-31.
Hix, Simon. 1998. “Dimensions and Alignments in European Union Politics: Cognitive Constraints and Partisan Responses.” European Journal of Political Research :forthcoming.
Hoffmann, S. 1966. “Ostinate or Obsolete? The Fate of the Nation State and the case of Western Europe.” Daedalus :826-915.
Hofstede, Geert. 1991. Cultures and organisations : software of the mind. London: MacGraw-Hill.
Hofstede, Geert. 1994. “Images of Europe.” The Netherlands' Journal of Social Sciences 30:63-82.
Hooghe, Liesbet. 1997. “Serving 'Europe' - Politcal Orientation of Senior Commission Officials.” European Integration online Papers (EIoP) 1: http://eiop.or.at/eiop/texte/1997-008a.htm.
Hooghe, Liesbet. 1998. “Euro-Socialists or Euro-Marketeers. Explaining Contending Orientations to European Capitalism Among Senior Commission Officials.” Pp. 32 in Paper prepared for the Bi-Annual Conference for Europeanists. Baltimore, USA.
Hooghe, Liesbet. 1999. “Supranational Activists or Intergovernemental Agents. Explaining the orientations of senior Commission officials towards European integration.” Comparative Political Studies 32:forthcoming.
Hooghe, Liesbet and Gary Marks. 1997. “The Making of a Polity : The Struggle Over European Integration.” European Integration online Papers (EIoP) 1: http://eiop.or.at/eiop/texte/1997-004a.htm.
Inglehart, R. 1977. The Silent Revolution : Changing Values and Political Styles among Western Publics. Princeton: Princeton University Press.
Inglehart, R. 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton University Press.
Janssen, Joseph I.H. 1991. “Postmaterialism, Cognitive Mobilization and Public Support for European Integration.” British Journal of Political Science 21:443-468.
Kerremans, Bart. 1996. “Do Institutions Make a Difference? Non-Institutionalism, Neo-Institutionalism, and the Logic of Common Decision-Making in the European Union.” Governance 2:217-240.
Kuhfeld, W.F., W.S. Sarle, and F.W. Young. 1985. “Methods of Generating Model Estimates in the PRINQUAL Macro.” Pp. 962-971 in SAS Users Group International Conference Proceedings : SUGI 10. Carry NC: SAS Institute Inc.
Lindberg, Leon N. and Stuart A. Scheingold. 1970. Europe's Would-Be Polity. London: Prentice-Hall.
March, James G. and Johan P. Olsen. 1984. “The New Institutionalism: Organizational Factors in Political Life.” American Political Science Review 78:734-749.
Marks, Gary. 1996. “An Actor-Centred Approach to Multi-Level Governance.” Journal of Regional and Federal Studies 6:20-38.
Marks, Gary, Liesbet Hooghe, and Kermit Blank. 1996. “European Integration from the 1980's : State-Centric v. Multilevel Governance.” Journal of Common Market Studies 34:341-378.
Martinotti, Guido and Sonia Stefanizzi. 1995. “European and the Nation State.” Pp. 163-189 in Public Opinion and Internationalized Governance, vol. 2, Beliefs in Government, edited by O. Niedermayer and R. Sinnott. New York: Oxford University Press.
Mazey, S. and J. Richardson. 1995. “Promiscuous Policymaking : The European Policy Style.” Pp. 337-359 in The State of the European Union, edited by C. Rhodes and S. Mazey. Boulder CO: Lynne Riener.
Metcalfe, Les. 1994. “International policy co-ordination and public management reform.” International Review of Administrative Sciences 60:271-290.
Moravcsik, Andrew. 1991. “Negotiating the Single European Act : National Interests and Conventional Statecraft in the European Community.” International Organisation 45:651-688.
Moravcsik, Andrew. 1993. “Preferences and Power in the European Community : A Liberal Intergovernemtalist Approach.” Journal of Common Market Studies 31:473-524.
Pag, Sabine. 1987. “The relation between the Commission and National Bureaucracies.” Pp. 443 - 496 in The European Administration, edited by S. Cassese: IISA.
Pollack, Mark A. 1996. “The New Institutionalism and EC Governance : The Promise and Limits of Institutional Analysis.” Governance : An International Journal of Policy and Administration 9:429-458.
Putnam, Robert D. 1988. “Diplomacy and domestic politics : the logic of two-level games.” International Organization 42:427-460.
Risse-Kappen, Thomas. 1996. “Exploring the Nature of the Beast : International Relations Theory and Comparative Policy Analysis Meet European Union.” Journal of Common Market Studies 34:53-80.
Risse-Kappen, Thomas. 1997a. “Bringing transnational relation back in: introduction.” Pp. 3-33 in Bringing Transnational Relations Back In. Non-State Actors, Domestic Structures and International Institutions, vol. 42, Cambridge Studies in International Relations, edited by T. Risse-Kappen. Cambridge: Cambridge University Press.
Risse-Kappen, Thomas. 1997b. “Structures of governance and transnational relations: what have we learned?” Pp. 280-313 in Bringing Transnational Relations Back In. Non-State Actors, Domestic Structures and International Institutions, vol. 42, Cambridge Studies in International Relations, edited by T. Risse-Kappen. Cambridge: Cambridge University Press.
Sinnott, Richard. 1995. “Policy, Subsidiarity, and Legitimacy.” Pp. 246-276 in Public Opinion and Internationalized Governance, vol. 2, Beliefs in Government, edited by O. Niedermayer and R. Sinnot. Oxford: Oxford University Press.
Taggart, Paul. 1998. “A touchstone of dissent: Euroscepticism in contemporary Western European party systems.” European Journal of Political Research 33:363-388.
Wessels, Bernhard. 1995a. “Development of Support : Diffusion or Demographic Replacement?” Pp. 105-136 in Public Opinion and Internationalized Governance, vol. 2, Beliefs in Government, edited by O. Niedermayer and R. Sinnott. New York: Oxford University Press.
Wessels, Bernhard. 1995b. “Support for Integration : Elite or Mass-Driven?” Pp. 137-162 in Public Opinion and Internationalized Governance, vol. 2, Beliefs in Government, edited by O. Niedermayer and R. Sinnott. New York: Oxford University Press.
Wessels, W. 1991. “The EC Council : the Community's decisionmaking center.” Pp. 133-154 in The new European Community. Decisionmaking and institutional change, edited by S. Hoffmann and R. O. Keohane. Boulder Co: Westview Press.
Westlake, Martin. 1995. The Council of the European Union. London: Catermill.
(*) This article was partly written while the author was a research assistant at the University of Antwerp. The research was made possible by a grant of the Fund for Scientific Research - Flanders (Fonds voor Wetenschappelijk Onderzoek - Vlaanderen, 073/1557) and was carried out under promotorship of Guido Dierickx. Ten interviewers of the ISPO (Interuniversitair Steunpunt Publiek Opinie Onderzoek) carried out the interviews. A draft of this paper was presented at the Third Pan-European Conference on International Relations in the panel on 'The European Union as a Negotiated Order' (Vienna, 16-19 September 1998, convenors: Ole Elgström and Mike Smith). The author wish to express his gratitude to Ann Carton of the ISPO for her help in doing the field work in Brussels, to Jan De Bock and Vincent Mertens de Wilmars of the Belgian Permanent Representation for the crucial information they provided me with and to Liesbet Hooghe, Bart Kerremans, Lewis Carrafiello, Morten Egeberg and Peter Bursens for the valuable comments and suggestions. I also wish to acknowledge valuable input from the two anonymous referees of the EIoP.
(1) The positioning of member-states on our scale corresponds for instance fairly well with the record of No-votes and abstentions listed in the European Voice, 15-21 October 1998.
(2) Neo-institutionalists reject the implicit notion of institutional efficiency in neo-functionalism (March and Olsen 1984). The latter posit that regional integration creates new policy problems (functional spill-over) which encourage the shifts of political loyalties to the supranational level (political spill-over) (Lindberg and Scheingold 1970). For neo-institutionalists institutions often lead to inefficient outcomes.
(3) Unfortunately we do not possess adequate data in order to examine the principal-agent dynamic Hooghe proposes to test.
(4) Since several of our hypotheses are about the potential importance of contextual factors we are tempted to separate individual attributes from social context in order to assess its importance for individual dispositions. Therefore the desaggregation of aggregate data leads to a conceptualisation of 'structure' or 'context' as an individual attribute of every official. This desaggregation has indeed some disadvantages with which we will deal later on.
(5) A negative impact of post-materialism on supranationalism could have been expected as well. Ingelhart himself observed a reversed relation between postmaterialism and support for integration for the Danish sample (Inglehart 1977). Ingelhart also suggested that postmaterialism could stimulate subnational identities. However, it remains unclear whether a subnational identity co-exists with a supranational identity. Also Dobratz found in analysing Greek Euro-barometer data (of 1980-1984) that support for integration in Greece declined as respondents showed more post-materialistic values (Dobratz 1993).
(6) Transgovernmental relations are conceived as a subset of transnational relations, which include multinational organisations, international non-governmental organisations and transgovernmental networks among state officials (Risse-Kappen 1997a). Transgovernmental relations concern networks among governmental actors deviating from the traditional image of inter-state relations between heads of state and governments. The concept covers governmental sub-units that handle relatively autonomously because national decisions, and or instructions are incomplete, imperfect or in some cases even absent. Thus, the actions of transgovernmental networks take place rather independent from national policy-making.
(7) In our sample it is even a very small minority that had previous experiences within the European Commission (Belgians = 3, non-Belgians = 3).
(8) The distinction North South is linked with issues such as deregulation and compensation of dislocation costs and structural policy-making (Garrett 1992), cultural heterogeneity (Egeberg 1996; Metcalfe 1994), variance in power resources (Moravcsik 1991) and differences in policy-making styles (Christoph 1993; Mazey and Richardson 1995).
(9) The distinction between net-receivers and net-payers could be an interesting independent variable as well. The disadvantage of this indicator is its very strong relation with the North-South dummy, an association that will create serious problems of collinearity within a multivariate analysis (r=0.68, p=0.00001). Consequently we propose to drop an explicit reference to the cleavage between net-payers and net-receivers and to employ the North-South distinction as a proxy for it.
(10) Also Hooghe situates her hypothesis concerning national elite orientations within a principal-agent logic, which implies in her study that national governments (the principals of the European Commission) will try to constrain the power of the European Commission (the agent of the national governments) (Hooghe 1999).
(11) About Italy Gabel and Palmer say: '… the positive image of EC membership as responsible for Italy's economic resurgence coupled with public frustration over political reform, has been credited with Italians' widespread support for a more federal European integration' (Gabel and Palmer 1995). Also Deflem and Pampel mention the Italian case: '… because of many scandals involving Italian politicians, Italian citizens may be expected to express less nationalistic and more pro-European sentiments' (Deflem and Pampel 1996).
(12) Unfortunately, we do not have more recent and similar quantitative data at our disposal. However, from extensive feedback we received from the Belgian Permanent Representation and the Belgian Foreign Office we believe that the picture we sketch in the paper is quite stable over time. Thus, recent, more qualitative, information confirms the quantitative analysis. Moreover, one has to keep in mind that the data concern political culture and attitudes. It is well known that this type of data is rather stable over time and that it is generally not directly connected with the content of the policy agenda of a particular moment. To put it differently, cultural features are quite enduring. This brings us to a third comment. Despite the fact that the data itself was collected in 1994, the paper deals with a problematique which is not very time bound, namely the consequences of socialisation of national officials, civil servants and politicians within a European polity.
(13) Parallel to these interviews 23 in-depth interviews were carried out. These interviews help to interpret the results of the quantitative analyses.
(14) From a semantic point of view Hooghe's items measure the same concept. These items are:
(15) All scales in this paper were also analysed with the help of non-linear, namely alternating least squares techniques (Kuhfeld, Sarle, and Young 1985). Components retained had to show an eigenvalue superior to 1 (the so-called Kaiser's criterion). For the further multivariate analysis we will deal with composite indices which reveal virtually exactly the same results as analyses on indices based on factor-scores.
(16) In order to makes averages comparable we employ standardised z-scores.
(17) Elsewhere we employed these items for the measurement of supranationalism (Beyers and Dierickx 1997). In fact these items are semantically related to what Sinnott calls 'attributed internationalisation', which supposes that attitudes are dependent upon the evaluation of policy problems (Sinnott 1995).
(18) An ad hoc hypothesis to explain the deviance of the German respondents could be that their responses were significantly more error prone than those of other nationalities. German responses behaviour was more affected by missing values and were less consistent.
(19) This finding runs parallel to Hooghe's multivariate analysis (Hooghe 1998).
(20) The correlation between both variables for the multinational sample is 0.46 (p=0.0001).