Gaustadalléen 30 (map)
ARENA Working Papers
Why Europeanisation Happens-
The Transformative Power of EU Committees
Norwegian Institute for Studies in Research
and Higher Education
Introduction  ![endif]>![if>
Processes of European integration have attracted major scholarly attention from within several disciplines in the post World War II period. One central research question has been how supranational the European Union (EU) is. A related question, which is put to the fore in this study, is to what extent EU decision-makers develop supranational loyalties (Haas 1958). This chapter studies the socialisation and re-socialisation of individuals, not states. Some scholars assume that EU decision-makers are re-socialised as far as their role and identity perceptions are concerned (e.g. Franklin and Scarrow 1999; Hayes-Renschaw and Wallace 1997:235; Joerges 1999:320; Laffan, O�Donnell and Smith 1999:87; Scully 2002; Weiler 1999:342). A common assertion in the literature is that an �esprit de corps� emerges within EU committees (Laffan 1998; Pag 1987), especially if the committee participants interact fairly frequently and intensively (e.g. Eriksen 2000:61; Haas 1958; Lewis 2000). Contrary to this, Pollack (1998) and Wessels (1998:227) argue that no loyalty transfers take place at the supranational level. Few empirical observations, however, are available that confirm or reject these assertions and assumptions. This study argues that both the above observations are partly correct when looking at national civil servants attending EU committees.
The study of European integration has developed from a study of EU institutions towards a study of the EU through institutional lenses.  ![endif]>![if> This chapter applies an institutional approach to supranationalism emphasising that national civil servants attending EU committees tend to supplement pre-existing allegiances with supranational role perceptions. According to an institutional approach EU committees are transformative institutions with respect to the roles of the committee participants. I argue that the length and intensity of attendance on EU committees accompanies the enactment of supranational allegiances and loyalties amongst the participants. Supranational allegiances denote identifying with EU institutions (cf. below).
Past studies demonstrate �the importance of national context and the capacity of national administrative traditions to modify, accommodate, internalise and, perhaps, even neutralise European pressures� (Goetz 2000:216). Processes of europeanisation do not replace or reject national administrative structures, cultures, rules and norms (Cowles, Coparaso and Risse 2001). This observation is strongly emphasised by Egeberg and by Beyers and Steensels in this volume. Similarly, the emergence of supranational allegiances amongst EU committee participants is not likely to replace pre-established national identities (Pollack 1998; Risse 2001). The main emphasis of our study, however, is to identify scope conditions that foster the emergence of supranational loyalties amongst EU committee participants. The two major scope conditions emphasised are:
The length of attendance on EU committees, and
The intensity of attendance on EU committees.
Operational measures of these scope conditions are indicated at pages 12 and 13.
The next section suggests an operational definition of supranationalism. Next, I present an institutional approach to explain why national civil servants attending EU committees tend to evoke supranational allegiances. The empirical analysis is based on survey and interview data from Danish and Swedish government officials who attend Council working parties (CWPs). CWPs are organised below the Council of Ministers and the COREPER (cf. Lewis� chapter). CWPs prepare dossiers for decision at the Minister level and consist of national bureaucrats together with Commission representatives and chaired by the Presidency of the EU. This study compares CWP participants coming from national ministries and agencies on the one hand and officials at the Permanent Representations in Brussels on the other. Our core assumption is that the latter tend to develop stronger supranational allegiances than the former due to more intensive and extensive participation within the CWPs. Moreover, Danish officials are expected to evoke supranational allegiances more strongly than their Swedish counterparts. This assumption derives from Danish officials having participated for longer periods of time in the CWPs than their Swedish colleagues. Together, figure 1 summarises the main empirical expectations of this chapter and the two central scope conditions emphasised:
Figure 1: A two-dimensional plot of empirical expectations
This chapter demonstrates that EU committees indeed are sites of socialisation and re-socialisation of national civil servants. The empirical analysis demonstrates that the transformative power of EU committees accompanies the evocation of supranational allegiances amongst those attending, especially amongst the permanent representatives and amongst the Danish officials. However, national officials and Swedish committee participants also evoke supranational allegiances. Secondly, the transformative power of EU committees is shown to be secondary to the influence generated by domestic government institutions. The effects generated by the length and intensity of attendance on EU committees are mediated and filtered by the primary institutional affiliations embedding the committee participants. This conclusion parallels the conclusions drawn by Beyers and Steensels, Egeberg and Lewis in this volume. Contrary to the assumption of Ernst Haas� (1958), supranational loyalties do not replace pre-existing national loyalties. They supplement them.
Civil servants are multiple and complex selves with different roles, identities and action modes. EU committee participants are indeed multiply embedded. This study argues that particular scope conditions activate particular repertoires within a set of roles and action modes and deactivate others. Supranational allegiances are activated particularly amongst senior EU committee participants who attend the committees with a high level of intensity. �Going native� in EU committees, however, do not imply �staying native� when the officials return to the national ministries and agencies. After their stay in Brussels, theses officials might re-activate national and sectoral roles and allegiances. The current study, however, emphasises that officials attending EU committees intensively are more likely to �stay native� than officials having only occasional trips to Brussels.
Towards an operational definition of supranationalism
The study of European integration can be divided into, at least, three scholarly traditions with different accounts of supranationalism � neo-functional, intergovernmental and institutional approaches. The latter approach will guide our empirical analysis. Early neo-functional approaches emphasised European integration as the horizontal integration in width and depth at the EU level of governance (e.g. Haas 1958). Empirical indications of integration in width were, amongst others, the number of issue areas covered by the Community (Lindberg and Scheingold 1970). Indicative of integration in depth was, among others, the usage of qualified majority voting in the Council of Ministers. According to neo-functionalists, supranationalism resembles a steadily increasing spill-over process across policy sectors as well as loyalty transfers from purely national institutions towards supranational institutions (Haas 1958; Saeter 1998). According to this perspective, supranational loyalties are detrimental to national allegiances.
Second, intergovernmental accounts on European integration have mainly studied this phenomenon as the horizontal co-operation between sovereign nation-states. Applying a two-level game approach, the EU integration process is perceived as the aggregate effect of bilateral negotiations amongst the EU member states (e.g. Milward 1992; Moravcsik 1998; Putnam 1988). According to this perspective supranationalism represents a �pooling of sovereignty� by sovereign national governments to EU institutions. EU committee participants have clear, written and mandatory instructions from their national principals. Accordingly, national allegiances tend to exceed supranational loyalties amongst EU committee participants.
A third and more recent analytical take on European integration, which underpins this study, pay particular attention to the vertical linkages between the EU level and the national level of governance. European integration is seen as resulting from blurring of these levels (e.g. Aspinwall and Schneider 2001; Bulmer 1997; Cowles, Coparaso and Risse 2001; Egeberg and Trondal 1999; Hanf and Soetendorp 1998; Held 1999:57; Lewis 2000; Rometsch and Wessels 1996). Accordingly, supranationalism denotes, amongst other things, the emergence of supranational role and identity perceptions amongst national government officials. This concept of supranationalism partially parallels the notion of loyalty transfers addressed by the neo-functionalist perspective. However, an institutional approach also pays attention to EU decision-makers as multiple selves with several partly conflicting identities, roles and cues for action. In the following sections an institutional approach to supranationalism is outlined and empirically tested. Blurring of governance levels is viewed synonymously to the evocation of supranational role perceptions amongst national civil servants attending EU committees.
National officials have their primary institutional affiliations and allegiances nationally (cf. Egeberg�s chapter). Participation on EU committees, however, can add new supranational allegiances to pre-existing ones (cf. the next section). Hence, the institutional perspective offered in this study resembles a middle ground between the neo-functional and the intergovernmental approaches with respect to the empirical predictions generated (cf. figure 1).
EU committees as sites of socialisation. Different empirical yardsticks can measure supranationalism. Institutionalist perspectives tend to view supranationalism as synonymous with the transformative power of EU institutions. One central assumption derived from institutional theories is that exposure to EU institutions, in general, and to EU committees, particularly, contributes to the subsequent emergence of supranational allegiances amongst the actors. EU committees are located at the very intersection of the national bureaucracy and the EU decision-making apparatus. EU committees represent the very �transmission belt� through which supranational allegiances are constructed (Christiansen and Kirchner 2000:22). Arguably, intensive and sustained interaction within EU committees accompanies the enactment of supranational allegiances amongst the national committee participants (cf. the next section). This study applies the following empirical proxies for supranational allegiances:
� Allegiances towards the CWPs attended, and![endif]>![if>
� Allegiances towards the EU as a whole.![endif]>![if>
Roles are defined in institutional terms in this study. That is, evoking roles means acquiring an ��organizational personality� rather distinct from his personality as an individual� (Simon 1957:278). Supranational roles resemble identifications with EU committees and with the EU as a whole. Moreover, roles and identities are considered in parallel because these properties constitute a continuum: Having an identity is a result of internalising the values and goals prescribed by the role (Barnett 1993:274). Moreover, by �role� we usually mean a set of expectations (norms or rules) that more or less specify the desired behaviour of the actor (March and Olsen 1989). Roles prescribe how one should act while identities in addition prescribe who one should be. Roles are linked to a feeling of belongingness to a community. �To the extent that organization members identify with their organization, they are willing to act spontaneously in its interest, without being told exactly what to do� (Mayntz 1999:83). Identification does not require that deviant desires or behavioural preferences be absent, only that internal (rather than external) sanctioning mechanisms are sufficiently effective to prevent deviant preferences from being brought into action. Hence, �the adoption of roles is central to the development of social identities� (Pratt 1998:196). According to social identity theory the adoption and enactment of roles are necessary means for producing identification (Pratt 1998). However, identities are more difficult to distinguish from the self than roles, and therefore also more difficult to alter in different situations. Identification with roles, however, need not come about because �it is relatively easy to set aside the reality of the secondary internalizations� (Berger and Luckman 1966:162). The enactment of supranational roles by EU committee participants need not accompany deep and lasting identifications with these roles. Hence, our primary focus is on the role perceptions of EU committee participants.
The concept of role rests on an analogy of the theatre where the actor is expected to perform according to a particular script (Stryker and Statham 1985:330). Consistent with an institutional perspective on organisations (cf. below), �[the theatre] consists of socially constructed players endowed with different capacities for action and parts to play� (Scott 1995:42). This study analyses civil servants as actors at two different theatres: domestic (ministries, agencies and permanent representations) and European (the CWPs). These officials are multiple selves with several potential roles and identities to enact (Barnett 1993; Elster 1986). The question thus becomes which script (role) should they enact in different plays? Additionally, is each official capable of separating different roles in different plays?
Roles are fairly stabile features of the self and hence relatively robust characteristics of actors embedded in organised communities (Heidar 1997:93; March and Olsen 1989). Consistent with a multiple concept of role actors can shift the attention towards different roles when changing play, albeit not always constructing qualitatively new roles for each new situation or play. Different plays often have points of resemblance triggering actors to evoke similar role perceptions. For example, national ministries and agencies have several institutional matches with different EU committees (Egeberg and Trondal 1999). Institutional compatibility can strengthen pre-existing and pre-established role perceptions amongst the committee participants. Institutional mis-match, however, is assumed to challenge pre-existing roles (see Egeberg�s contribution to this volume). This argument derives from the institutional approach to supranationalism outlined below which emphasises the multiple institutional embeddedness of EU committee participants.
Why study the roles activated by individual decision-makers? First, they provide cognitive, moral and normative systems of orientation and self-reference. This study emphasises role conceptions, as perceived by the officials themselves. We study role perceptions of civil servants as they �exist in the minds of [civil servants]� (Saalfeld and M�ller 1997:9). Second, they provide the actors with shared systems of meaning. Third, they influence the framing of action: �What [people] do and how they do it depends upon how they see themselves and their world, and this in turn depends upon the concepts through which they see� (Pitkin 1972:1). Hence, role perceptions provide �conceptions of reality, standards of assessment, affective ties, and endowments, and thereby with a capacity for purposeful action� (March and Olsen 1995:30). Subsequently, studying the roles of EU committee participants is important for understanding European integration through transformative processes at the level of individual decision-makers (Christiansen, J�rgensen and Wiener 2001).
Mechanisms of socialisation: An institutional approach to supranationalism
The former section has briefly sketched the neo-functional and the intergovernmental approaches as well as suggested an institutional perspective as a middle ground between these. This middle ground is not founded solely on theoretical mechanisms but also on empirical predictions. The neo-functional and the intergovernmental perspectives expect us to observe the emergence of either supranational allegiances or national roles among EU committee participants, respectively. The institutional approach, however, moves away from either/or towards both/and theorising. According to an institutional perspective both supranational and national allegiances might be evoked by EU committee participants, however, under different institutional conditions.
New-institutional perspectives in organisational analyses present several mechanisms of socialisation and transformational change (DiMaggio and Powell 1991; March and Olsen 1989 and 1995; Peters 1999; Scott 1987). One common denominator integrating this plurality is the emphasis put on contextualised, endogenous decision behaviour, identity and senses of belonging. People are perceived as homo politicus as much as homo economicus. Attention is directed towards the way different institutional contexts mould behaviour and roles differently. Institutions not only constrain these elements, they also contribute to the initial construction of them (Checkel 2001).
A cognitive organisation theory perspective has developed within social psychology and introduced to organisational theory largely by Simon (1957), March and Simon (1958), and Cyert and March (1963). Information-based and knowledge-based models in the explanation of political dynamics are thus not new. Hence, the current interest in the cognitive dimension of politics has been characterised as more of a rediscovery than of absolute novelty (Pollack 1998; Radaelli 1999:757). The hypotheses suggested in this study on the socialisation of EU committee participants, draws heavily on a cognitive perspective on organisations. The onthology of methodological individualism underlying the cognitive perspective downplays the role of social interaction and puts emphasis on the organisational arrangements in which such interaction occurs. Hence, our focus is directed to the organisational embeddedness of social interaction among EU committee participants.
A cognitive perspective emphasises that individuals� role perceptions and preferences are endogenous and thus possible to construct and reconstruct. A cognitive organisational perspective stresses how exposure towards particular organisational structures contributes to the enactment of particular role perceptions and codes of conduct among individual decision-makers. Organisations are mechanisms of simplification with respect to information processing (Simon 1957). The underlying assumption is that of bounded rationality. The possibility for individuals to attend everything simultaneously is impossible. Attention is a scarce resource. To be able to move beyond individual bounded rationality a cognitive perspective pictures organisational structures as mechanisms for coupling and de-coupling actors, problems, solutions, consequences, roles and institutional allegiances. Organisational borders are buffers to attention, thereby biasing the information exposed to each decision-maker (March and Olsen 1995; Tenbrunsel et al. 1996).
For organisational designers, one way of reducing information-overload is to carve up organisations horizontally and vertically thereby creating buffers against particular actors, certain information, considerations and stimuli (Gulick 1937; Schattschneider 1960). �Cognitive structures simplify when there is too much, and they thus allow the perceiver to reduce an enormously complex environment to a manageable number of meaningful categories� (Markus and Zajonc 1984:143). Within a cognitive approach organisational structures represent cognitive buffers to attention and information. �Because of the limits of human intellective capacities in comparison with the complexities of the problems that individuals and organizations face, rational behaviour calls for simplified models that capture the main features of a problem without capturing all its complexities� (March and Simon 1981:148). Organisational structures render it possible to decompose complex tasks into sub-tasks that can be carried out within relatively independent units of governance. According to cognitive psychology social identity processes are ultimately governed by the individual need for uncertainty reduction as regards their �perceptions, attitudes, feelings, and behaviours�� (Hogg and Terry 2000:124). Organisation structures contribute to the development of �cognitive short cuts� for individual decision-makers (Johnson 1987:45). These shortcuts contribute to the creation of cognitive categories and simplified representations of world phenomena to these individuals. As a result, these phenomena are taken for granted as �the way we do these things� (Scott 1995:44). Organizational boundaries affect role perceptions because these properties simplify cognitive search processes and reduce cognitive uncertainty (Johnson 1987; March 1994; Simon 1957:288). Cognitive scripts provide �guidelines for sense-making and choosing meaningful action� (Scott 1995:44).
Organisational members are collections of identities, roles and modes of behaviour. They are multiple selves (Elster 1986). Studies demonstrate that national and supranational identities correlate positively (Licata 2000; Risse 2001). However, several of the chapters in this volume show that supranational roles are indeed secondary to national and sectoral roles (cf. the contributions of Beyers and Steensels, Egeberg, and Lewis). Moreover, the evocation of multiple roles might blur the distinctiveness of each role (cf. Lewis� chapter in this volume). I basically argue, however, that officials evoking multiple roles manage to separate one role from the others, thus keeping each role basically distinct.
Departing from this assumption, this study analyses the conditions under which supranational roles gain particular strength. Our primary argument is that, when �members of one polity serve as participants in the political processes of another� (Rosenau 1969:46), as when domestic officials participate on CWPs, the length and intensity of participation on EU committees affects the extent to which supranational role perceptions are evoked amongst the participants (Hooghe 1999). Hence, length and intensity are the scope conditions emphasised in this chapter. Apart from being formal members of EU committees, protracted and intensive interaction and participation within these committees is conducive to the evocation of supranational role perceptions amongst the committee members. Parallel to this argument, Haas (1958) assumed that participants become �locked in� and socialised by the sheer intensity of interaction. �The interactive character of decision making extends over time so that the development of beliefs, rules and expectations in one organization is intertwined with their development in others� (March 1999:29)
Our argument rests upon socialisation theory which emphasises a positive relationship between the intensity of participation within a collective group and the extent to which the members of this group take the world for granted (Meyer and Rowan 1991), become victims of �group think� (Janis 1982; t�Hart et al. 1997), or develop particular �community methods� (Lewis 2000; Smith 1998). According to contact theory, the length and intensity of contacts affects the attitudes of the actors positively (Pettigrew 1998). Socialisation processes are dynamic whereby the actors come to identify with their government roles, with their government institutions as well as with the government machinery as a whole. If national government officials are to be socialised into supranational actors, the duration and intensity of interaction at the supranational level arguably matters. Socialisation also represents interactive processes between the �socialisators� and the �socialisees�. Moreover, socialisation processes are uni-directional in the sense that the �socialisator� educates, indoctrinates, teaches, persuades and diffuses his norms, beliefs and rules to the �socialisee� (Berger and Luckmann 1966:150; Kerr 1973; Pendergast 1976; Smith 1992:58).  ![endif]>![if> Hence, protracted and intensive exposure towards CWPs (the �socialisators�) increases the likelihood that the participants (the �socialisees) enact supranational roles. Similar to neo-functionalist analyses, the cognitive institutional approach views the evocation of supranational allegiances to be �a function of the duration of the socialization impact� (Niemann 1998:437 - emphasis added; McDonald 1998:51; Pettigrew 1998). Accordingly, �[m]any roles are learned through playing the roles�� (Stryker 1980:63).  ![endif]>![if>
Contrary to the socialisation thesis, one might assume that protracted exposure towards certain institutions teach actors how to de-couple talk and action - thus keeping the role and identity perceptions of the actors largely unchanged and unaffected by ways of presenting them (March 1984). For example, EU committee participants may be inclined to adopt �Euro-talk� picturing oneself as �Europeanized�, i.e. using �Euro-jargon�, whilst at the same time acting in accordance with national expectations and obligations. Combining pre-established national modes of acting and feeling and applying standarized and �Europeanized� language, civil servants may be able to satisfy inconsistent expectations and demands. �Euro-talk� might thus reflect �the norms geared exclusively for talk� more than the norms of action and practice (Brunsson 1998: 267). In the current study, however, no efforts are made studying the relationships between talk and action. Role and identity perceptions are studied separately from decision behaviour. However, it might be questioned whether the role and identity perceptions evoked by officials are myths or reality to them (Brunsson 1989). Solid theoretical solution to this puzzle has not been suggested in the literature, nor are we trying to solve this puzzle here. When are processes of socialisation and re-socialisation occuring and when do individuals �talk to talk� (Moravcsik 2001:237)?
According to the above arguments, national civil servants are likely to identify with EU institutions as a result of �daily reinforcement� and intensive exposure towards new stimuli and information at the EU level (Lodge 1978:241; Kerr 1973). Consequently, officials devoting much time and energy attending CWPs are likely to take on supranational roles. That is, they identify with the CWPs attended, with the Council of Ministers, or with the EU as a whole. National civil servants become increasingly EU-minded and loyal to EU policies and politics. They become increasingly europeanised.
Intensive and prolonged participation on EU committees represents the two scope conditions suggested in this chapter. The following four operational measures of our two main scope conditions are applied in the empirical analysis below:
� Prolonged and sustained participation on EU committees,![endif]>![if>
� attendance on many EU committees and many committee meetings,![endif]>![if>
� being an active member of the committees by giving frequent oral presentations during committee meetings, and![endif]>![if>
� having informal contacts outside the formal committee meetings.![endif]>![if>
The first of the above empirical proxies measures the length of participation on EU committees � that is, the time span from their first EU committee meeting up to 1999 when our empirical study was conducted (cf. the section on data and methodology). The other three variables listed above measure the intensity of attendance on EU committees � that is, the degree of participation at any point in time, both past and present. In example, some of the Danish officials studied made their first appearance in the EU committees in the 1970s but have seldom attended these committees with any degree of intensity. Other officials, both Swedish and Danish, attended their first EU committee meeting in the 1990s and have participated in a whole lot of meetings since then.
CWPs are collegial arrangements of a non-permanent nature. Similar to Beyers and Steensels, and Egeberg in this volume, CWPs are composed largely of �part-timers� whose primary institutional affiliations lie elsewhere. The socialisation potential is assumed to be, and is also empirically shown to be, weaker within non-permanent collegial organisations than within permanent hierarchical organisations. This is empirically revealed within the European Parliament (Bowler and Farrell 1995; Katz 1997; Scully 1999), within the American Congress (Fenno 1962), and within EU committees (Egeberg 1999; Trondal 2001; Trondal and Veggeland 2003). Collegial organisations are composed of members who are pre-socialised and �pre-packed� before attending the collegium. Notwithstanding the pre-socialisation of civil servants, the re-socialization potential of the collegium which they attend is assumed to be strengthened if they attend the committee often, if he or she is a senior participant, if the same participants meet regularly, and if each colleague generally devote a major amount of time participating within the collegial setting (Checkel 2001; Dierickx and Beyers 1999; Lewis 2000; Trondal and Veggeland 2003). Thus, senior EU committee participants are likely to take on supranational allegiances more extensively than officials who are newcomers at the EU arena (Beyers 1998). The length and intensity of attendance is generally assumed to blur the organisational borders between the collegium and the �core-organisation�; in this study blurring the borderlines between the EU level and the domestic central administrative systems. This argument, however, does not neglect the multiple institutional embeddedness of government actors and the pre-socialisation of these actors at the national level (cf. the contributions of Egeberg in this volume). Our main emphasis in this chapter, however, is to highlight the impact of EU committee attendance.
The empirical record: data and methodology
The unit of analysis in this study is Danish and Swedish civil servants who participate on CWPs. Commission expert committees and comitology committees are excluded from this analysis due to methodological considerations.  ![endif]>![if> In order to control for the potential effects of different EU committees only the CWPs are selected to this study. The CWPs are selected because they enable us to compare officials from domestic ministries and agencies on the one hand and officials at the permanent representations to the EU on the other. Few permanent representatives attend Commission expert committees or comitology committees (Trondal 2001).
The choice of individual civil servants as our research units is based on two rationales: First, officials are those that develop feelings of belonging, who are exposed to role expectations and role prescriptions, and who, ultimately, act. Second, many domestic officials are exposed to impulses from different social and societal contexts. They are �full-timers� within domestic government institutions while at the same time being �part-timers� within the CWPs.  ![endif]>![if> Their primary institutional affiliations are the domestic ministries, agencies, and the permanent representations in Brussels. The EU committees are only secondary institutional affiliations to these officials (cf. Egeberg�s chapter). The central question put forward in this study is to what extent intensive and lengthy secondary affiliations to EU committees accompany the enactment of new supranational role perceptions amongst the committee participants. Control for the role of pre-socialisation at the national level is conducted empirically in the next section.
To test the relationships between the length and intensity of participation on EU committees and supranational allegiances, research units have been selected that allow for sufficient variation on these variables. Two samples of respondents were systematically selected that included national civil servants that had attended EU committees for different length of time and with different levels of intensity. Based on these samples, one survey was conducted from summer 1998 to spring 1999. This survey originally covered Danish, Swedish and Norwegian ministry and agency officials attending Commission expert committees and CWPs. The current chapter, however, study only Danish and Swedish officials because the Norwegian officials do not attend the CWPs. This survey is based on a standardised questionnaire, which was sent to each civil servant by mail. The response rate is 73 per cent, giving 116 respondents. Additionally, a triangulation strategy was adopted to increase the validity of the analysis. Hence, 22 face-to-face interviews were conducted amongst some of the Danish and Swedish officials. Those officials selected for interview were drawn from the above survey sample. The interviews were used as a follow-up to the empirical findings in the survey. A semi-structured interview-guide laid the basis for the interviews. In order to control for potential effects of policy sector, an empirical circumscription of the study is done by way of including only two policy sectors: the environment sector and the field of health and safety at the workplace. However, no efforts are made analysing potential effects of different sector affiliations among the officials.
Most Swedish civil servants have participated on EU committees for relatively short periods of time because Sweden became an EEA member  ![endif]>![if> in 1994 and a full-fledged EU-member in 1995. Danish government officials, on the contrary, have had access to the CWPs since 1973. The above interview and survey data on Swedish and Danish officials allows for testing the extent to which the length of attendance on EU committees affects the enactment of supranational loyalties amongst the participants.
Second, as to account for the intensity dimension, officials from national ministries and agencies are assumed to attend CWPs with less intensity than officials at the Permanent Missions to the EU (cf. figure 1). A survey was conducted at the Swedish and Danish Permanent Representations to the EU from fall 1998 to spring 1999. A similar questionnaire as presented to the ministry and agency officials was used at the Permanent Representations. The response rate in this survey is 55 per cent, giving 41 respondents. This lower response rate partly reflects a high workload at the Permanent Missions at the time of study and partly a somewhat weaker effort to increase the response rate on behalf of the author. Permanent representatives are expected to participate more intensively on CWPs than ministry and agency personnel. Hence, supranational allegiances are likely to be evoked more extensively amongst the former than amongst the latter.  ![endif]>![if> The study of permanent representatives does no sectoral circumscription similar to that in the national sample. Selecting permanent representatives from different policy fields as well as officials from the diplomatic realm increase the sheer number of observations in our study (N). By increasing N, the number of statistical techniques available is also larger.
Applying survey data and interview data from �national officials� and permanent representatives provides ample possibilities to test the relationships between attendance on EU committees and the emergence of supranational allegiances amongst the participants. These data also warrant testing the extent to which supranational allegiances reflect the length and intensity of attendance. However, a perhaps more ideal data set would have included panel data. Panel data warrants studying the same individuals in time sequences. Hence, panel data could more adequately demonstrate identity and role changes across time. Time series and iterated interviews would most likely have increased the dynamic element in the analysis. Moreover, actual observations of committee meetings could also have moved the analysis closer to the question of social interaction within the EU committees.
The data presented above is based on a systematic selection procedure. Based on this procedure, no statistical or empirical generalisations are possible to do. The only possible road to empirical generalisations is by referring to other empirical studies � in this volume or elsewhere - which supports or rejects my findings. Moreover, theoretical generalisations are done by referring to our theoretical universe presented above as well as in chapter 1. Our empirical observations are guided by fairly general theoretical arguments and might be used as a partial empirical testing of these arguments.
The salience of supranational allegiances
Civil servants are complex selves. They have several institutional affiliations, representational roles and cues for action. National civil servants attending EU committees have their primary institutional affiliations at the national level of governance. They are employed in domestic ministries, agencies and permanent representations, in different formal positions, and have various educational backgrounds. Moreover, the CWPs are mainly organised according to a territorial principle of organisation, and the participants are often guided by governmental instructions indicating the positions to be pursued during committee meetings (cf. Egeberg in this volume; Schaefer et al. 2000; Trondal 2001). Hence, national officials attending CWPs are likely to put strong emphasis on national allegiances. As such, our research design resembles a least likely design. Even under these conditions, however, our study emphasises that supranational allegiances are likely to supplement pre-established allegiances amongst senior EU committee participants.
Table 1 demonstrates the relative mix of different institutional allegiances evoked by national officials attending CWPs. This table makes a distinction between officials coming from national ministries and agencies on the one hand, and officials at the Permanent Representations in Brussels on the other.
Table 1 Distributions of allegiances to domestic government institutions, policy sectors, professional backgrounds, and supranational institutions (%).a
a) The dependent variables combine value 1 and 2 on the following five-point scale: very great extent (1), fairly great extent (2), both/and (3), fairly small extent (4), very small extent (5).![endif]>![if>
b) For the permanent representatives this variable measures allegiances to their �own� Permanent Representation.![endif]>![if>
Table 1 shows that CWP participants have a strong national role orientation. Most CWP participants feel allegiances to their own government institution and to their �own� national government (cf. Egeberg�s chapter in this volume). However, due to permanent representatives having only temporary posts at the Brussels delegations, allegiances towards this institution is lower than the corresponding allegiances amongst the national officials. Second, table 1 demonstrates that the second most important allegiance reported is sector. Finally, supranational allegiances are weaker than both national and sectoral allegiances. Similar to the observations made by Beyers and Steensels, Egeberg, and Lewis in this volume, supranational identities are indeed secondary. These observations are also supported in the interview data. One Swedish national official argued that,
�I feel strongest allegiance to Sweden, but I develop a certain loyalty to the committee. Still, this loyalty never exceeds the loyalty to [my national institution]� (author�s translation). Similarly, one Danish national official claimed that �I have the strongest national identity, but this [identification] should be in accordance with the principles of the [EU] Treaty� (author�s translation).
Hence, when attending EU committees supranational allegiances are more weakly enacted than pre-existing role perceptions.
�It is hard to change ways of acting and thinking� (Swedish national official � author�s translation).
Table 1 also shows that permanent representatives tend to have stronger supranational allegiances than national officials. One important observation in this respect is that permanent representatives attending CWPs feel stronger allegiances towards the committees attended than to their educational backgrounds. The officials studied are heavily pre-socialised through national (and possibly international) educational institutions. Despite these kinds of pre-socialisation, the officials are re-socialised fairly strongly within the EU committees. Yet, this re-socialisation process does not replace pre-established national and sectoral allegiances.
The two following tables reveals the distributions of supranational allegiances when controlling for (i) the intensity of attendance on CWPs (table 2), and the length of attendance on CWPs (table 3).
Table 2 Distributions of supranational allegiances amongst Danish and Swedish officials, controlled for the intensity of attendance on CWPs (national officials vs. permanent representatives) (%).a
a) The dependent variables combine value 1 and 2 on the following five-point scale: very great extent (1), fairly great extent (2), both/and (3), fairly small extent (4), very small extent (5).![endif]>![if>
Table 3 Distributions of supranational allegiances amongst national officials and permanent representatives, controlled for the length of attendance on CWPs (Danish vs. Swedish officials) (%).a
a) The dependent variables combine value 1 and 2 on the following five-point scale: very great extent (1), fairly great extent (2), both/and (3), fairly small extent (4), very small extent (5).![endif]>![if>
Both table 2 and table 3 demonstrates that the intensity of attendance on the CWPs have an independent effect on supranational allegiances. Only weak support is lent to the assumed effects of the length of participation. A more adequate multivariate analysis of the relative explanatory effect of the length and intensity of attendance on CWPs is provided in tables 5 and 6 (see below).
The next question is to what extent these supranational allegiances reflect re-socialisation processes within the CWPs or processes of pre-socialisation prior to entering the CWPs. Are national civil servants generally supranational before attending EU committees or do they become increasingly europeanised during their �stay� at the EU committees? How can we be certain that supranationm socialisation al loyalties stem frodynamics at the EU level and not out of the sheer self-selection of civil servants with pre-established supranational loyalties (Pollack 1998:27)? One way of testing this is to study how the participants are selected to the EU committees. Our data reveal that the majority of the sampled EU committee participants (80 per cent of the national officials and 95 per cent of the permanent representatives) are routinely invited to attend meetings in the CWPs. Hence, only a minority of committee participants has been subject to self-selection to these committees. However, one could assume that officials with pre-established supranational loyalties are over-represented in the national civil service at large, especially at the Permanent Missions in Brussels. Still, it is likely that recruitment to these institutions generally is based on merit rather than on particular institutional loyalties. For most of the national civil servants, participation in EU committees represents an integral part of their otherwise complex portfolio. Therefore, supranational allegiances can hardly reflect processes of self-selection to the CWPs.
Why supranational allegiances emerge
According to the institutional arguments presented above, officials devoting much time and energy in EU committees are more likely to develop new supranational roles than officials devoting less time and energy in the EU committees. Hence, supranational allegiances are likely to reflect daily exposure towards supranational institutions. This section also reveals the length and intensity to which national civil servants and permanent representatives actually attend CWPs and the extent to which Danish and Swedish officials differ in this respect. Secondly, this section demonstrates through regression analysis to what extent the length and intensity of attendance on the CWPs accompanies supranational allegiances.
First, table 4 shows the intensity to which national officials and permanent representatives attend CWPs.
Table 4 Distributions of intensity of attendance on CWPs (average numbers and %).
a) Values 1 and 2 are combined on the following five-point scale: very great extent (1), fairly great extent (2), both/and (3), fairly little extent (4), very little extent (5).![endif]>![if>
b) This variable and the next variable combine values 1 and 2 on the following five-point scale: very often (1), fairly often (2), both/and (3), fairly seldom (4), very seldom (5).![endif]>![if>
Due to Denmark being a more senior EU member than Sweden, Danish officials have, on average, attended the CWPs for longer periods of time than the Swedish officials. In our sample, 100 per cent of the Swedish officials attended the CWPs for the first time in 1994 or later � reflecting the EEA agreement between the EU and the EFTA countries. Most of the sampled Danish officials made their first appearances in the CWPs prior to 1994. Consequently, most of the Danish officials have participated for longer periods of time in the CWPs compared to the Swedish officials.
Table 4 shows that national civil servants devote a fairly great amount of time participating on CWPs, provide frequent oral presentations during committee meetings and have extensive informal contacts outside the formal committee meetings. Table 4 also demonstrates that officials at the Permanent epresentations attend the CWPs much more intensively than the national officials (cf. Lewis� chapter in this volume). For example, officials at the Permanent Representations attend a greater number of CWPs than national officials. National officials have on average attended 3 CWPs. National officials also attended on average 6 CWP meetings during the last year (1998). Permanent representatives, in contrast, have on average attended 19 CWPs. These officials also attended on average 82 CWP meetings during the last year (1998). Together these observations give ample indications that officials at the Permanent Representations in Brussels attend the CWPs much more intensively than national officials from the capitals. These observations also support the observations made in tables 1 and 3: permanent representatives evoke supranational allegiances more strongly than national officials due to their intensity of attendance on the CWPs.
In the following, the above observations are used in multiple regression analyses to test to what extent the length and intensity of participation on the CWPs accompanies supranational allegiances amongst the participants. These regression analyses apply two dependent variables: (i) allegiances towards the CWPs (table 5), and (ii) allegiances towards the EU as a whole (table 6). The independent variables included in the regressions are those presented in table 4, supplemented by one additional independent variable discussed above: the length to which officials have attended the CWPs. Together, the two regression models analysed below (tables 5 and 6) enable us to measure the relative explanatory power of the two central scope conditions in our chapter: the length and the intensity of attending EU committees. In both tables several operational measures of the intensity variable is applied (see above for a complete list of these measures). Tables 5 and 6, however, only present those regression coefficients that are significant at the 95 per cent and the 99 per cent levels. Due to low N in the analyses (cf. the above tables) significant relationships are hard to find. Therefore, those regression coefficients that are .40, or stronger, are also presented even if they are not significant. This procedure is sound because significance does not relate to any theoretical universe in this study. Due to the fact that the samples are based on systematic selection procedures, significance tests demonstrate, at best, the robustness of the relationships studied.  ![endif]>![if>
Table 5 analyses to what extent officials who have participated for long periods of time and with a high degree of intensity on the CWPs come to feel allegiances towards these committees.
Table 5 Factors related to allegiances towards the CWPs attended. Regression coefficients (beta).a
*) p � .05��������� **) p � .01��������������������������������������������������������������� R2 = .31��������������������������������� R2 = .40
a) The dependent variables have the following values: very great extent (1), fairly great extent (2), both/and (3), fairly small extent (4), very small extent (5).![endif]>![if>
b) This variable has values that correspond to the actual number of committees attended. Attendance on many committees is given a high value, while attendance on few committees is given a low value. For example, attendance on one committee is given the value of 1.![endif]>![if>
Table 5 shows that only two independent variables relate significantly with the dependent variable. These variables relate, however, only to one of our scope conditions: the intensity of attendance on the CWPs. No empirical support is provided for the assumption that length of participation on EU committees fosters supranational allegiances. Notwithstanding few significant observations in table 5, the explained variance (R2) is fairly high in each analysis. The two significant observations presented above support the proposed pattern. Permanent representatives who devote much time participating on CWPs tend to feel allegiances towards these committees (.50*), and national officials who have many informal face-to-face contacts with fellow committee participants tend to feel allegiance towards the CWPs (.59*). Hence, senior committee participants who engage in informal networking with other committee participants tend to evoke supranational allegiances. Lewis (2000) and Trondal (2001) make similar observations. These findings are also supported in our interview data. One Danish national official argued that,
�The essential happens in the breaks [between the formal meetings] (author�s translation). Similarly, one Swedish national official observed that, �we have frequent contacts between the meetings, rather informal personal contacts. This results in a certain allegiance to the committee and to the individuals who attend. I almost know the committee participants better than my colleagues back home. We turn into a club� (author�s translation).
Finally, table 6 demonstrates similar empirical patterns with respect to allegiances towards the EU as a whole.
Table 5 Factors related to allegiances towards the EU as a whole. Regression coefficients (beta).a
*) p � .05��������� **) p � .01��������������������������������������������������������������� R2 = .13����������������������������� R2 = .32
c) The dependent variables have the following values: very great extent (1), fairly great extent (2), both/and (3), fairly small extent (4), very small extent (5).![endif]>![if>
d) This variable has values that correspond to the actual number of committees attended. Attendance on many committees is given a high value, while attendance on few committees is given a low value. For example, attendance on one committee is given the value of 1.![endif]>![if>
Similar to table 5, table 6 demonstrates that the intensity of attendance on the CWPs is conducive to the emergence of supranational allegiances among the participants. More precisely, table 6 reveals that officials having attended many CWPs also tend to feel allegiance towards the EU as a whole (.42 and .60*). One Swedish national official argued that,
�a feeling of participation in the EU � as an organisation � develops� (author�s translation).
Parallel to table 5, the explained variance (R2) is higher in the regression analysis on the permanent representatives than in the analysis on the national officials. Moreover, the observations presented in tables 1 and 3 are supported in tables 5 and 6. Permanent representatives attend the CWPs with a higher degree of intensiby than the national officials, accompanying stronger supranational allegiances amongst the former. However, no solid empirical support is provided for the assumption that length of participation on the CWPs is conducive to supranational allegiances. Hence, only one of our two scope conditions is empirically supported. This might partly reflect the fact that our empirical operationalisation of the length variable is less rich than the operationalisations of the intensity variable. However, the observed pattern is also likely to reflect the fact that the intensity of attendance on EU committees is a stronger scope of re-socialisation than the sheer length of participation.
This chapter has demonstrated that EU committees indeed are sites of socialisation and re-socialisation of national civil servants. Processes of re-socialisation towards supranational allegiances are shown to reflect one scope condition: the intensity of attendance. National civil servants attending EU committees fairly intensively tend to evoke supranational allegiances more strongly than officials who attend EU committees with less intensity. This study also shows that officials at the Brussels-based Permanent Representations attend CWPs more intensively than national officials from the capitals. This difference accompanies stronger supranational allegiances amongst permanent representatives than amongst national officials. Finally, the regression analyses demonstrate that length of attendance is a weaker scope condition than intensity of attendance as to explain supranational allegiances amongst EU committee participants.
This study, however, also shows that the transformative power of EU committees is secondary to the influence generated by domestic government institutions (cf. table 1). The effects generated by the length and intensity of attendance on EU committees are mediated and filtered by the primary institutional affiliations embedding the committee participants. This conclusion parallels the conclusions drawn by Beyers and Steensels, Egeberg, and Lewis in this volume.
Contrary to Haas� (1958) assumption that the emergence of supranational loyalties is detrimental to pre-established national allegiances, the current study demonstrates that supranational allegiances and national/sectoral allegiances might co-exist (cf. table 1). Civil servants are multiple and complex selves with different roles, identities and action modes. This study has argued that particular scope conditions activate particular repertoires within a set of roles and action modes and deactivate others. Supranational allegiances are activated particularly amongst senior EU committee participants who attend the committees with a high level of intensity. �Going native� in EU committees, however, do not imply �staying native� when the officials return to the national ministries and agencies. After their stay in Brussels, theses officials might re-activate national and sectoral roles and allegiances. The current study has, however, emphasised that officials attending EU committees intensively are more likely to �stay native� than officials having only occasional trips to Brussels.
Aspinwall, M. D. and G. Schneider (2001) �Institutional research on the European Union: mapping the field�, in M. D. Aspinwall and G. Schneider (eds.) The Rules of Integration: The Institutionalist Approach to European Studies. Manchester: Manchester University Press.
Barnett, M. (1993) �Institutions, Roles, and Disorder: The Case of the Arab States System�, International Studies Quarterly, Vol. 37: 271-296.
Berger, P. L. and T. Luckmann (1966) The Social Construction of Reality. A Treatise in the Sociology of Knowledge. New York: Penguin Books.
Beyers, J. (1998) �How Supranational is Supranationalism? National and European Socialization of Negotiators in the Council of Ministers�, Acta Politica, Vol. 33, No. 3: 378-408.
Bowler, S. and D. M. Farrell (1995) �The Organizing of the European Parliament: Committees, Specialization and Co-ordination�, British Journal of Political Science, Vol. 25:219-243.
Brunsson, N. (1989) The organization of hypocrisy: talk, decisions and actions in organizations. Chichester: Wiley.
Brunsson, N. (1998) �Homogeneity and Hetrogeneity in Organizational Forms as the Result of Cropping-up Processes�, in N. Brunsson and J. P Olsen (eds.) Organizing Organizations. Oslo: Fagbokforlaget.
Bulmer, S. J. (1997) �New Institutionalism, The Single Market and EU Governance�, ARENA Working paper, No. 25/1997.
Checkel, J. T. (2001) �Why Comply? Social Learning and European Identity Change�, International Organization, Vol. 55, No. 3: 553-588.
Christiansen, T. and E. Kirchner (eds.) (2000) Europe in Change. Committee governance in the European Union. Manchester: Manchester University Press.
Cowles, M. G., J. Caporaso, and T. Risse (2001) Transforming Europe: Europeanization and Domestic Change. Ithaca: Cornell University Press.
Cyert, R. M. and J. G. March (1963) A behavioral theory of the firm. Englewood Cliffs, New York: Prentice-Hall.
Demmke, C. (1998) �The Secret Life of Comitology or the Role of Public Officials in EC Environment Policy�, EIPASCOPE, No. 3: 14-23.
Dierickx, G. and J. Beyers (1999) �Belgian Civil Servants in the European Union. A tale of two cultures�, West European Politics, Vol. 22, No. 3: 198-222.
DiMaggio, P. J. and W. W. Powell (1991) �Introduction�, in W. W. Powell and P. J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis. Chicago: The University of Chicago Press.
Egeberg, M. (1999) �Transcending Intergovernmentalism? Identity and Role Perceptions of National Officials in EU decision-making, Journal of European Public Policy, Vol. 6, No. 3: 456.474.
Egeberg, M. and J. Trondal (1999) �Differentiated integration in Europe: the case of the EEA country Norway�, Journal of Common Market Studies, Vol. 37, No. 1: 133-142.
Elster, J. (1986) The Multiple Self. Cambridge: Cambridge University Press.
Eriksen, E. O. (2000) �Deliberative supranationalism in the EU�, in E. O. Eriksen and J. E. Fossum (eds.) Democracy in the European Union. Integration through Deliberation? London: Routledge.
Fenno, R. F. (1962) �The House Appropriations Committee as a Political System: The Problem of Integration�, The Americal Political Science Review, Vol. LVI: 310-324.
Franklin, M. N. and S. E. Scarrow (1999) �Making Europeans? The Socializing Power of the European Parliament�, in R. S. Katz and B. Wessels (eds.) The European Parliament, the National Parliaments, and European Integration. Oxford: Oxford University Press.
Goetz, K. H. (2000) �European Integration and National Executives: A Cause in Search of an Effect?�, West European Politics, Special Issue on Europeanised Politics? European Integration and National Political Systems (eds. K. H. Goetz and S. Hix), Vol. 23, No. 4: 211-231.
Gulick, L. (1937) �Notes on The Theory of Organizations. With Special References to Government in the United States�, in L. Gulick and D. Urwick (eds.) Papers on the Science of Administration. New York: Institute of Public Administration, Columbia University.
Haas, E. (1958) The Uniting of Europe. Stanford: Stanford University Press.
Hanf, K. and B. Soetendorp (eds.) (1998) Adapting to European Integration. Small States and the European Union. London: Longman.
Hart, P., E. K. Stern and B. Sundelius (eds.) (1997) Beyond Groupthink. Political Group Dynamics and Foreign Policy-making. Michigan: The University of Michigan Press.
Hayes-Renshaw, F. and H. Wallace (1997) The Council of Ministers. New York: St. Martin�s Press.
Heidar, K. (1997) �Roles, Structures and Behaviour: Norwegian Parliamentarians in the Nineties�, The Journal of Legislative Studies � special issue on Members of Parliament in Western Europe: Roles and Behaviour, Vol. 3, No. 1: 91-108.
Held, D. et al. (1999) Global Transformations: Politics, Economics and Culture. Cambridge. Polity Press.
Hogg, M. A. And D. J. Terry (2000) �Social Identity and Self-Categorization Processes in Organizational Contexts�, Academy of Management Review, Vol. 25, No. 1: 121-140.
Hooghe, L. (1999) �Supranational Activists or Intergovernmental Agents? Explaining the Orientations of Senior Commission Officials Towards European Integration�, Comparative Political Studies, Vol 32, No. 4: 435-453.
Institut f�r Europ�ische Politik (1987) Comitology � Characteristics, Performance and Options. Preliminary Final Report. Bonn.
Janis, I. L. (1982) Groupthink. Psychological Studies of Policy Decisions and Fiascoes. Boston: Houghton Mifflin Company.
Joerges, C. (1999) �Bureaucratic Nightmare, Technocratic Regime and the Dream of Good Transnational Governance�, in C. Joerges and E. Vos (eds.) EU Committees: Social Regulation, Law and Politics. Oxford: Hart Publishing.
Johnson, G. (1987) Strategic Change and the Management Process. New York: Bail Blackwell.
Katz, R. S. (1997) �Representational roles�, European Journal of Political Science, Vol. 32: 211-226.
Kerr, H. H. Jr. (1973) �Changing Attitudes through International Participation: European Parlamentarians and Integration�, International Organization, Vol. 27, No. 1: 45-83.
Laffan, B. (1998) �The European Union: a distinctive model of internationalization�, Journal of European Public Policy, Vol. 5, No. 2: 235-253.
Laffan, B., R. O�Donnell and M. Smith (1999) Europe�s Experimental Union. Rethinking Integration. London: Routledge.
Lewis, J. (2000) �The methods of community in EU decision-making and administrative Rivalry in the Council�s intrastructure�, Journal of European Public Policy, Vol. 7, No. 2: 261-289.
Licata, L. (2000) �National and European Identities: complementary or antagonistic?, 1999-2000 European Forum, IDNET Conference Europeanization and Multiple Identities, Conference paper, EUR/67.
Lindberg, L. and S. Scheingold (1970) Europe�s Would-Be Polity. Patterns of Change in the European Community. Englewood Cliffs, New York: Prentice-Hall.
Lodge, J. (1978) �Loyalty and the EEC: The Limitation of the Functionalist Approach�, Political Studies, Vol. XXVI, No. 2: 232-248.
McDonald, M. (1998) �Identities in the European Commission�, in N. Nugent (ed.) At the Heart of the Union. Studies of the European Commission. Houndmills: MacMillan Press.
March, J. G. (1984) �How We Talk and How We Act: Administrative Theory and Administrative Life�, in T. J. Sergiovanni and J. E. Corbally (eds.) Leadership and Organizational Cultures. Urbana: University of Illinois Press.
March, J. G. (1994) A Primer on Decision Making. How Decisions Happen. New York: The Free Press.
March, J. G. (1999) The Pursuit of Organizational Intelligence. Oxford: Blackwell.
March, J. G. and J. P. Olsen (1989) Rediscovering Institutions. The Organizational Basis of Politics. New York: The Free Press.
March, J. G. and J. P. Olsen (1995) Democratic Governance. New York: The Free Press.
March, J. G. and H. A. Simon (1958) Organizations. New York: Wiley.
March, J. G. and H. A. Simon (1981) �Decision-Making Theory�, in O. Grusky and G. A. Miller (eds.) The Sociology of Organizations. Basic Studies. Second Edition. New York: The Free Press.
Markus, H. and R. B. Zajonc (1985) �The Cognitive Perspective in Social Psychology�, in G. Lindzey and E. Aronsen (eds.) Handbook of Social Psychology. Volume I: Theory and Method. New York: Random House.
Mayntz, R. (1999) �Organizations, Agents and Representatives�, in M. Egeberg and P. L�greid (eds.) Organizing Political Institutions. Essays for Johan P. Olsen. Oslo: Scandinavian University Press.
Meyer, J. W. and B. Rowan (1991) �Institutionalised Organizations: Formal Structure as Myth and Ceremony�, in W. W. Powell and P. J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis. Chicago: The University of Chicago Press.
Milward, A. S. (1992) The European Rescue of the Nation-State. London: Routledge.
Moravcsik, A. (1998) The Choice for Europe. Social Purpose and State Power from Messina to Maastricht. Ithaca: Cornell University Press.
Moravcsik, A. (2001) �Bringing Constructivist Integration Theory Out of the Clouds: Has it Landet Yet?�, European Union Politics, Forum Section: A Constructivist Research Program in EU Studies?, Vol. 2, No. 2: 226-240.
Niemann, A. (1998) �The PHARE programme and the concept of spillover: neofunctionalism in the making�, Journal of European Public Policy, Vol. 5, No. 3: 428-446.
Pag, S. (1987) �The relations between the Commission and national bureaucracies�, in S. Cassese (ed.) The European Administration. Brussels and Maastricht: IIAS and EIPA.
Pendergast, W. R. (1976) �Roles and attitudes of French and Italian delegates to the European Community�, International Organization, Vol. 30, No. 4: 669-677.
Peters, B. G. (1999) Institutional Theory in Political Science. The �New Institutionalism�. London: Pinter.
Pettigrew, T. F. (1998) �Intergroup Contact Theory�, Annual Review of Psychology, Vol. 49: 65-85.
Pitkin, H. F. (1972) The Concept of Representation. Berkeley: University of California Press.
Pratt, M. G. (1998) �To Be or Not to Be: Central Questions in Organizational Identification�, in D. A. Whetten and P. C. Godfrey (eds.) Identity in Organizations. Building Theory Through Conversations. Thousand Oaks: SAGE.
Putnam, R. D. (1988) �Diplomacy and Domestic Politics: The Logic of Two-level-Games�, International Organization, Vol. 42, No. 3: 427-460.
Radaelli, C. M. (1999) �The public policy of the European Union: Wither politics of Expertice?�, Journal of European Public Policy, Vol. 6, No. 5: 757-774.
Risse, T. (2001) �A European Identity? Europeanization and the Evolution of Nation-State Identities�, in M. G. Cowles, J. Caporaso and T. Risse (eds.) Transforming Europe. Europeanization and Domestic Change. Ithaca: Cornell University Press.
Rometsch, D. and W. Wessels (eds.) (1996) The European Union and member states. Towards institutional fusion? Manchester: Manchester University Press.
Rosenau, J. N. (1969) �Towards the Study of National-International Linkages�, in Rosenau (ed.) Linkage Politics. Essays on the Convergence of National and International Systems. New York: The Free Press.
Saalfeld, T. and W. C. M�ller (1997) �Roles in Legislative Studies: A Theoretical Introduction�, The Journal of Legislative Studies � special Issue on Members of Parliament in Western Europe: Roles and Behaviour, Vol. 3, No. 1: 1-16.
Saeter, M. (1998) Comprehensive Neofunctionalism. NUPI Report. Oslo.
Sandholtz, W. and A. Stone Sweet (eds.) (1998) European Integration and Supranational Governance. Oxford: Oxford University Press.
Schaefer, G. F., M. Egeberg, S. Korez and J. Trondal (2000) �The Experience of Member State Officials in EU Committees�, EIPASCOPE, No. 3: 29-35.
Schattschneider, E. E. (1960) The Semi-Sovereign People. New York: Holt, Rinehart & Winston.
Scott, W. R. (1987) �The adolescence of Institutional Theory�, Administrative Science Quarterly, Vol. 32: 493-511.
Scott, W. R. (1995) Institutions and Organizations. Thousand Oaks: SAGE Publications.
Scully, R. M. (1999) �Between Nation, Party and Identity: A Study of European Parliamentarians�, paper presented at the ECSA conference, Pittsburgh 2-5 June, 1999.
Scully, R. M. (2002) �Going Native? Institutional and Partisan Loyalties in the European Parliament�, in B. Steunenberg and J. Thomassen (eds.) The European Parliament on the Move: Towards Parliamentary Democracy in Europe. Rowan and Littlefield (forthcoming).
Sen, A. (1999) Reason before Identity. The Romanes Lecture for 1998. New York: Oxford University Press.
Shore, C. (2000) Building Europe. The Cultural Politics of European Integration. London: Routledge.
Simon, H. (1957) Administrative Behavior. New York: Macmillan.
Smith, A. D. (1992) �National identity, and the idea of European unity�, International Affairs, Vol. 68, No. 1: 55-76.
Smith, M. E. (1998) �Rules, Transgovernmentalism, and the Expansion of European Political Cooperation�, in W. Sandholtz and A. Stone Sweet (eds.) European Integration and Supranational Governance. Oxford: Oxford University Press.
Stryker, S. (1980) Symbolic Interactionism. Menlo Park, California: The Benjamin/ Cummings Publishing Company.
Stryker, S. and A. Statham (1985) �Symbolic Interaction and Role Theory�, in G. Lindzey and E. Aronson (eds.) Handbook of Social Psychology. Volume I. Theory and Method. New York: Random House.
Tenbrunsel, A. E. et al. (1996) �Cognitions in Organizations�, in S. R. Clegg, C. Hardy and W. R.. Nord (eds.) Handbook of Organizational Studies. London: SAGE.
Trondal, J. (2001) Administrative Integration across Levels of Governance. Integration through Participation in EU Committees. PhD dissertation, University of Oslo and ARENA Report, No. 7.
Trondal, J and F. Veggeland (2003) �Access, voice and loyalty. The representation of domestic civil servants in EU Committees�, Journal of European Public Policy, Vol. 10, No. 1 (forthcoming).
Van Schendelen, M.P.C.M. (1996) �EC Committees: Influence Counts More Than Legal Powers�, in R. H. Pedler and G. F. Sch�fer (eds.) Shaping European Law and Policy. The Role of Committees and Comitology in the Political Process. Maastricht: EIPA.
Weiler, J. H. H. (1999) �Epilogue: �Comitology� as Revolution � Infranationalism, Constitutionalism and Democracy�, in C. Joerges and E. Vos (eds.) EU Committees: Social Regulation, Law and Politics. Oxford: Hart Publishing.
Wessels, W. (1998) �Comitology: fusion in action. Politico-administrative trends in the EU system�, Journal of European Public Policy, Vol. 5, No. 2: 209-234.
 ![endif]>![if> This study is jointly financed by the ARENA programme (The Norwegian Research Council) and the Robert Schuman Centre for Advanced studies, EUI, Florence (The European Forum). An earlier version of this chapter was presented at the workshop �Socialization and Identity Change�, 16-17 June 2000 Oslo, and at the workshop �International Institutions and Socialization in the New Europe�, 18-19 May 2001 Florence. I would like to thank the participants at these workshops for valuable and valid comments. Special thanks go to Jan Beyers, Jeffrey T. Checkel, Morten Egeberg, Iain Johnston, Judith Kelly, Jeffrey Lewis, Mark Pollack, Frank Schimmelfennig, Thomas Risse and Michael Zuern for valuable and insightful comments.
 ![endif]>![if> Thanks to John Erik Fossum at ARENA for this point.
 ![endif]>![if> The role of agency, however, is not ruled out of the model: Which role and mode of action that is evoked is ultimately a matter of choice, however, largely biased by cognitive limitations and mental maps. Rational choice is conditioned by institutional contexts that provide cues for selecting certain roles and modes of acting above others (Sen 1999). Hence, agency is contextualised and embedded within organisation structures.�
 ![endif]>![if> The sheer size of societal communities, organisations or collegial arrangements may affect the role perceptions enacted by individuals, and this dimension may also to some extent condition the impact of the intensity dimension. CWPs normally convene from 20 to over 40 people. Consequently, the potential for intimacy and close bargaining and arguing is provided by the sheer size of these committees and groups. This intimacy may provide for an �esprit de corps� to emerge amongst the participants. However, the committees covered by this study are of approximately the same size. Hence, no hypotheses are generated on the basis of the size variable that might explain variance in our dependent variables.
 ![endif]>![if> Empirical studies show that officials participating on Commission expert committees and comitology committees have difficulties separating these committees with respect to their formal status (Institut f�r Europ�ische Politik 1987:81; Van Schendelen 1996). �This mixture of working group and comitology committee sometimes makes it very difficult for national civil servants to know when they have to act as representative of a Member State within a Comitology committee and when as an independent national expert� (Demmke 1998:17). Excluding both of these committees from the analysis helps reducing the likelihood of mixing different committees, albeit not completely excluding this possibility.
 ![endif]>![if> The distinction between �full-timers� and �part-timers� is not solely a question of the time and energy devoted by officials towards different institutional contexts. In addition, this distinction is based upon formal organisational characteristics. In that respect, domestic officials who attend EU committees should be conceived as �full-timers� within their domestic ministries or agencies and only �part-timers� at the EU-level of governance.
 ![endif]>![if> The European Economic Area.
 ![endif]>![if> A more extensive presentation and discussion of the data and methodology is provided in Trondal 2001.
 ![endif]>![if> Diagnostics of collinearity between the independent variables analysed in tables 5 and 6 unveil no indications of extreme multicollinearity. Thus, the independent variables seem to have independent causal impact on the two dependent variables.