XLIII ECPR Joint Sessions of Workshops. Conceptualising Interest Groups: An Addition to the 1980s

Transcription

1 XLIII ECPR Joint Sessions of Workshops University of Warsaw (Poland) 29th March 2nd April 2015 Workshop: Conceptualising and Comparing Interest Groups and Interest Group Systems Chairs: Claudius Wagemann and Andrea Pritoni Conceptualising Interest Groups: An Addition to the 1980s Andrea Pritoni 1 and Claudius Wagemann 2 ABSTRACT: The so-called OBI project ( Organization of Business Interests ), which Schmitter and Streeck carried out in the early 1980s, provided subsequent scholars with a framework for the analysis of interest groups. One major contribution in this respect was the development of two different sets of opposite associational logics: on the one hand, interest groups can emphasize the logic of influence or the logic of membership ; on the other hand, interest groups are also challenged by finding a position between the logic of effective implementation and the logic of goal formation. This paper takes up this framework systematically and conceptualizes and operationalizes the ideal types which are derived from these logics. It is discussed with the help of set theory, in particular fuzzy sets, in how far the oppositional character of these logics can and must be overcome. KEYWORDS: Interest Groups Quali-quantitative Methods Conceptual Analysis Set Theory 1 Andrea Pritoni is post-doctoral fellow in the Department of Political and Social Sciences at the University of Bologna. His research interests are mainly oriented towards the fields of comparative politics (in particular with respect to parliamentary governments), interest groups, and policy analysis (with a particular focus on the banking and insurance sector). He has just published a book on the Italian Association of Banks and several articles on Italian interest groups. 2 Claudius Wagemann is full professor of social science methodology in the Department of Political Science at the Goethe University Frankfurt. With regard to interest group research, he published a book with Routledge on Private Interest Governments and several articles and book contributions. He has also widely published on qualitative comparative methods, most notably QCA and fuzzy sets. Furthermore, his research agenda includes right-wing extremism from a social movement perspective, political parties, and the Quality of Democracy.

2 1 Introduction Sometimes, social science research areas are shaped by contributions which divide the discussion in before and after. With regard to interest group research, this is the case with Schmitter s (1974) seminal article on neo-corporatism. This article has mainly become known as a caesura with regard to reasoning about interest groups. Above all, it added another view to the dominant pluralist perspective (most famously, Truman 1951), namely neo-corporatism, and thus placed a new emphasis on the insight that interest groups very often play a crucial role in the policy-making process. Neo-corporatism quickly diffused into the research community (Czada 1994), and its empirical manifestation in real-world policymaking was confirmed, despite periodical ups and downs (Schmitter 1989; Schmitter & Grote 1997). Additionally, this theoretical advancement was followed by an intense empirical research activity which culminated in the late 1970s / early 1980s in the so-called OBI ( Organization of Business Interests ) project. The initial reflections on the design of this major project (Schmitter & Streeck 1999 [1981]) did not only become of practical use for the numerous (mainly junior) researchers involved in OBI, but have also led to many more fundamental insights about interest groups 3 which have survived the actual empirical project. Next to the development of guidelines for organizational analysis which differentiate the organizational properties of interest domains, structures, resources, and policy outputs (Schmitter & Streeck 1999 [1981]: 45ff.), two sets of associational logics, namely the logic of influence and the logic of membership, and the logic of goal formation and the logic of effective implementation have become famous. The latter have then also been integrated with the notion of various organizational properties and types of goods and finally being enlarged to a typology of four ideal types (see below and Schmitter & Streeck 1999 [1981]: 21) have become famous. What followed was that an impressive amount of empirical research on interest groups has been produced (Hojnacki et al. 2012; Baumgartner & Bunea 2014). However, conceptual analysis has not been at the core of the activities; indeed, conceptual, methodological and disciplinary barriers seem to have rendered the accumulation of knowledge in interest group research quite difficult (Beyers et al. 2008: 1103) 4. Nevertheless, scholars still continue to devote little attention to the analysis of the fundamental concepts on which IG scholarship is based. This, of course, also has notable (negative) effects on organizational analysis. The understanding of organizations oftentimes seems limited to the questions of organizational resources which interest groups are able to spend in order to have a role in the policymaking process (Crombez 2002; Eising 2007), as well as to those of internal democracy 3 While Schmitter and Streeck (1999 [1981]) had designed their original ideas for business interest associations, some of their input is so generic that we refer to interest groups in general in this text, instead of to only the business sector as the original idea was. 4 Not even the same definition of interest group is taken for granted in the literature (Baroni et al. 2014). A key distinction can be made between a behavioural definition (Salisbury 1984; Baumgartner et al. 2009) and an organisational definition (Jordan & Greenan 2012; Binderkrantz et al. 2014). In the first case, groups are defined based on their observable, policy-related activities; in the second case, the interest group term is reserved only for membership associations. 2

3 (Halpin 2006; Binderkrantz 2009), whereas (almost) nothing has been said with respect to their intra-organizational structures to which the famous four logics allude. This paper departs from the general shortcoming of conceptual analysis in interest group research and focuses on the four logics as famous representatives of path-breaking work from more than 30 years ago. More precisely, it will present an attempt to conceptualize and operationalize the two dichotomies (logic of influence vs. logic of membership and logic of goal formation vs. logic of effective implementation) via indicators for the four ideal types (government, firm, club, movement) which can be derived from their interrelation. This will open the question whether there is really a conceptual opposition between pairs of ideal types which seems to be a consequence from earlier conceptualization. This will also make it possible to link the abovementioned great amount of recent empirical contributions to the interest group literature with a more sophisticated conceptual analysis. With regard to this, we propose twelve (three for each ideal type) possible indicators accounting for how much an interest group is government-like, firm-like, club-like or movement-like; during the course of our argumentation, we recur to fuzzy set theory in order to ascertain whether or not on the base of these indicators real-world interest groups follow the oppositional basis which is implicit in the associative logics. We proceed as follows: in an opening section (par. 2), we introduce Schmitter s and Streeck s (1999 [1981]) path-breaking contribution. Then (par. 3), we recur to the concept of property space developed by Lazarsfeld (1937) in order to question the orthogonality of the four logics. In the next section (par. 4), we propose potential indicators for all the four ideal types developed by Schmitter and Streeck. This is followed (par. 5) by an overview on how and why pairs of logics (as well as pairs of ideal types) are not true opposites to one another, making use of operations known from fuzzy set theory. Finally (par. 6), we make some concluding remarks and suggestions for future research. 2 Schmitter s and Streeck s Four Logics: Looking Back to the 1980s As mentioned, the project design for the OBI project (Schmitter & Streeck 1999 [1981]) had an enormous conceptual impact on the subsequent study of interest groups. Not only did the paper provide hands-on advice on different dimensions of organizational analysis (interest domains, structures, policy outputs, and resources), but has mainly become famous through the introduction of a terminology which was represented as a typology of structural components of interest groups. At the heart of this brilliant use of typology, a graph was developed (figure 1). 3

4 Figure 1: Schmitter s and Streeck s (1999 [1981]: 21) scheme on interest group logics As can be seen, this graph mainly develops around two axes which depict organizational logics: organizations can follow a logic of influence, recognizing the characteristics of their state interlocutors, or a logic of membership, oriented towards the characteristics of the organizational members. Furthermore, the graph visualizes the difference between the logic of goal formation and the logic of effective implementation. This distinction goes back to Child et al. s (1973) earlier work on trade unions in which these two logics were named representative rationality and administrative rationality, respectively (Schmitter & Streeck 1999 [1981]: 19). Admittedly, although this is rather an anecdotal observation than the result of careful empirical analysis, the former pair of logics has received greater scholarly attention than the latter. In any case, these two pairs of logics complement each other: while the logic of influence and the logic of membership indicate the nature of the dominant exchange relationship, the logic of goal formation and the logic of effective implementation focus on the internal organization. As a consequence, the four logically possible combinations of the two axes result in different structural ideal types to which associations can become similar or, ideally, equal. Schmitter and Streeck (1999 [1981]: 21) identify the ideal type of a firm which provides services through selective goods (logic of membership and logic of effective 4

5 implementation); of a government which sells compliance by means of monopoly goods and keeps control over its members (logic of influence and logic of effective implementation); of a movement which exerts pressure for public goods, representing its members (logic of influence and logic of goal formation); and of a club which creates consensus around solidaristic goods and has mainly participatory properties (logic of membership and logic of goal formation). This typology clearly went beyond the purely illustrative purpose of many 2x2 tables, using real-world labels for the ideal types. At the same time, it took the notion of ideal types seriously, since, when getting too close to the ideal type, an association runs the risk to change into something different (namely, a government, a movement, a club, or a firm). Moreover, this set of logics proved its analytical power by developing further into spin-off discussions, such as the literature on selective goods (Olson 1971 [1965]; Moe 1980), which has influenced interest group research from a conceptual point of view, or the research area on Private Interest Governments (Streeck & Schmitter 1985), which became substantially path-breaking. In any case, Schmitter and Streeck (1999 [1981]: 21) had produced a masterpiece which clearly organized the interest group world along two dimensions which were both dichotomous and which led to metaphorical descriptions of interest groups through realworld organizational forms which were formulated as ideal types. A basic underlying principle of the graph seems to be an either or perspective on the two dichotomous logics, although the two authors were not explicit on this. An interest group could either be oriented more towards the logic of membership or to the logic of influence; and either more to the logic of effective implementation or to the logic of goal formation. A gain towards one side had to be compensated through a loss at the other side. In short, Schmitter and Streeck produced a very useful framework for analysis which was well received in the scholarly discussion, and which was implicitly strongly marked through the dichotomous (and oppositional) understanding of the underlying logics shaping associational structures. In the following, we will challenge some elements of the picture, looking at it from new perspectives on conceptual analysis. 3 Schmitter s and Streeck s Typology as a Property Space In a certain sense, the picture proposed by Schmitter and Streeck (1999 [1981]: 21) reflects what Lazarsfeld (1937) has called a property space. 5 The central idea of this is that every case can be univocally located in a property space with regard to its specific manifestation of the constituting dimensions. In other words, an association underlining very much the logic of influence (and thus disregarding the logic of membership), while being indifferent with regard to the vertical axis of the logics, would have to be placed somewhere at mid-height in the right part of the graph. 5 Note that Lazarsfeld presented his ideas with the example of three dimensions; therefore, he uses the metaphor of a space. Schmitter s and Streeck s two-dimensional graph would imply a property plain. 5

6 Property spaces à la Lazarsfeld have recently been re-discovered in yet another literature, namely in the methodological innovation of Qualitative Comparative Analysis (QCA) (Ragin 1987; 2000; 2008; Schneider & Wagemann 2012). During a preparatory phase for the execution of the algorithm, this technique foresees so-called fuzzy values to be attributed to cases. Fuzzy values indicate in how far a given case is a member of a set; e.g., a membership value of 0.4 for an association to be a member in the set of successful associations indicates that the association is rather out of than in the set of successful associations, but that it is close to the cross-over point which differentiates between successful and other associations. 6 This step of assigning fuzzy values to cases (i.e., of assessing set membership scores) is also called calibration (Ragin 2008: 71ff.; Schneider & Wagemann 2012: 32ff.). Fuzzy scales are characterized by combining differences in kind and differences in degree (Schneider & Wagemann 2012: 14). While the widely used dichotomies also capture differences in kind, they fail in establishing differences in degree; 7 in contrast, ordinal or continuous scales as they are known from statistical analysis do indicate differences in degree, but often fail with regard to differences in kind, because the concepts they describe are not dichotomous in nature (or the dichotomy is not made explicit). So, it does not come as a surprise that scholars working on modes of concept formation in the social sciences became aware of the notion of fuzzy sets and implemented this idea into concept building strategies (Goertz 2005). Most concepts used in the social sciences are implicitly fuzzy, and making use of various properties of set theory could eventually lead to better and more sophisticated concepts. Without any doubt, Schmitter s and Streeck s pairwise logics can be considered fuzzy scales: they establish differences in kind, expressed in dichotomies which contradict one another, and they differentiate between various intensities of the manifestation of the logic. Also, the notion of ideal types which is derived corresponds to the ideal typical thinking of set theory and QCA where, at a certain point of the analytical process, cases are located in property spaces, depending on their distance to the ideal types the dichotomy represents (Schneider & Wagemann 2012: 103). Schmitter s and Streeck s idea is even more sophisticated than what usually happens with set theory, namely they creatively combine two fuzzy scales and derive four ideal types from these scales. If, however, the two dimensions leading to the four ideal types represent fuzzy scales, then the four ideal types are also in complementary relations: the ideal type of an interest group as a government is the logical and empirical opposite to the ideal type of an interest group as a club ; similarly, the ideal types of movements and firms become true opposites to one another. In other words, there are more oppositions (and thus fuzzy scales) in the graph (Figure 1) than just two pairwise logics. Pairs of ideal types also represent fuzzy scales, with the ideal types being the endpoints of the scales. 6 More precisely, fuzzy values (between 0 and 1) indicate the membership of cases in sets. If a case is a full member of a set (i.e., if the set characteristics describe the case sufficiently perfectly), then the value of 1 is attributed; if the case is a full non-member (i.e., the case does not reflect the set characteristics at all), then the value of 0 is given, with all the various gradations between 0 and 1. 7 If no differences in degree are established, the QCA literature calls the sets crisp sets (Schneider & Wagemann 2012: 24ff.). 6

7 4 Flesh to the Bones: Indicators for Ideal Types As we claimed in the introduction, while Schmitter s and Streeck s conceptual framework influenced to a large extent interest group scholarship to date, it is nevertheless weakened by some important lacunae: one of the main shortcomings is that there have never been systematic operationalizations of the two pairs of logics as well as of the two pairs of ideal types stemming from their interrelation. By focusing on ideal types rather than on logics, it becomes easier to place real-world associations somewhere in the picture, not least since the ideal types explicitly serve at connecting conceptual thinking with real-world objects. While finding indicators for the four logics can be regarded as a difficult, if not impossible, task to do, things become easier when we try to find indicators for the four ideal types, since they stand for organizational forms which we know quite well. Let us imagine the following list of indicators would be developed for deciding to which ideal types an association belongs: Table 1: Potential indicators for associational ideal types Ideal Type Specialized Activity Empirical Indicators Firm Providing services Legal and/or fiscal advice for associates Financing supply to associates Training courses for members Government Selling compliance No. of granted licenses and/or certifications to associates No. of negative sanctions inflicted on associates Solution to litigation among associates Movement Exerting pressure No. of public hearings No. of press releases No. of strikes Club Creating consensus No. of organized seminars and/or conferences No. of Assembly meetings Newsletter for associates As becomes evident from Figure 1, Schmitter and Streeck claimed that: i) firm-like interest groups provide services through selective goods; ii) government-like interest groups sell compliance by means of monopoly goods; iii) movement-like interest groups exert pressure for public goods; and iv) club-like interest groups create consensus around solidaristic goals. Therefore, we have used these types of goods in order to derive indicators for organizational types. Selective goods are private goods that an interest group produces exclusively for its members, in order to differentiate and eventually protect compliant members so that they are withheld from non-members or members violating their formal obligations towards the 7

8 interest group (Schmitter & Streeck 1999 [1981], 88ff.). The important point about this kind of goods is that they are individually usable by members; in addition, they can be considered as positive sanctions (rewards, incentives) for joining an association as well as negative sanctions to the extent that they can be withdrawn from non-compliant members. We decided to differentiate these goods into three categories: legal/fiscal services; economic services; educational services. As a consequence, the indicators we selected in order to account for how much a real-world interest group is similar to a firm, are the following: providing legal and/or fiscal advice for members; financing supply to members; developing training courses for members. As far as selling compliance is concerned, we can claim that all interest groups negotiating with policy makers or other interest groups on behalf of their members have a crucial interest in being able to act as an unitary actor (Schmitter & Streeck 1999 [1981], 90ff.). Whether or not (as well as how much) they are able to do so depends on their capability of inflicting negative sanctions 8 and/or of rewarding their associates with positive sanctions. 9 Therefore, we chose the following three indicators in order to account for how much a given interest group approaches the government-like ideal type: the number of licenses and/or certifications it grants to members; the number of negative sanctions it inflicts to them; whether or not it is able to give solution to litigations among members. Maybe the first and foremost activity of any interest group is to exert pressure on policy makers in order to get representation and to obtain favorable policy decisions. To this respect, interest groups which want being influential in the policy process specialize in lobbying activities. In the literature, a crucial distinction is made between direct lobbying, i.e. direct relationships with politicians and bureaucrats, and indirect lobbying, namely appealing to media or mobilizing its own membership (Grant 1989; Maloney et al. 1994). Therefore, we chose the number of public hearings as an indicator for direct lobbying, whereas the number of press releases as well as the number of strikes are indicators for indirect lobbying. Solidaristic goods are generally considered as to be connected to sociability, participation in collective activities, prestige, connections, formation of a collective identity and the like (Schmitter & Streeck 1999 [1981], 86ff.). The important points about this kind of goods are that they are collective goods namely they involve the interest group in its entirety as well as they stand for building up, reinforcing, and sustaining the identification of the members with their association. To this respect, we thus decided to select the number of organized seminars and/or conferences, the number of Assembly meetings (and, more in general, how many times organizational bodies are called to round up) and whether or not interest groups provide newsletters to members, as indicators of how much a real-world association is similar to a club. 5 From Fuzzy Sets to Profiles of Interest Groups 8 Negative sanctions are generally punishments inflicted by an interest group, or by the State on behalf of an interest group, upon non-compliant members. 9 Positive sanctions are usually rewards the interest group gives to its members for compliance with its decisions. 8

9 It is easy to imagine that these indicators as well as any enlarged version of our brief list could enable researchers to assign fuzzy values to single associations, with regard to their being similar to a club, a government, a movement, or a firm. In the end, every association would be characterized by four fuzzy values which describe its belongingness to the four ideal types. However, while this strategy sounds straightforward and could ultimately lead to a precise positioning of individual IGs somewhere in the graph, two observations disturb the picture: first, the conceptual opposition of firms and movements, and between governments and clubs, respectively, is terminologically not so clear as the one between the logics of influence and membership. More precisely, the question emerges to which shared dimension the pairs of ideal types belong: what are the features with regard to which firms and movements, or governments and clubs, respectively, are so different to one another? The second observation, however, is even more important and disturbing and can also be seen as the underlying reason for the first observation of terminological confusion. Empirically, it can happen (perhaps we can even dare to say that it is even probable) that interest groups behave both as firms do and as the (seemingly) opposed movements do. There is no reason why one and the same association cannot be active in legal advice and engage in unconventional forms of participation, such as strikes. In other words, while being attributed high fuzzy values for being a firm, there is no logical reason why an association cannot also be attributed high fuzzy values for being a movement. As we see, the nature of real-world complex associations can allow for behaving as a government as well as a club, or as a firm as well as a movement. At this point, a mathematical discussion of this observation is on order: fuzzy sets follow the rules of fuzzy negation (Schneider & Wagemann 2012: 47). The membership value of a case in the complement (i.e., the negation) of a set is always calculated by subtracting the membership value in the original set from 1: ~A = 1 A If a case has a fuzzy membership value of 0.8 in a given set, then it will automatically have a fuzzy membership value of 0.2 in the complement of the set. This rule also means that the fuzzy values of a case in a set and in the complementary set will sum up to 1. However, as introduced before, with the indicators we have chosen, it is not excluded that fuzzy membership values of two seemingly opposite ideal types will be both above 0.5 (i.e., that the association is similar to both ideal types of opposed ends), and thus the sum of the values is also above 1. In a similar logic, the sum could also be below 1, if both membership values of the opposed ideal types are below 0.5, or if the weak manifestation of one type is not compensated through an equally strong manifestation of the other type. In fact, it is an illusion to think that, when working with real-world indicators, the sum of the resulting indicators would be exactly 1. What does this mean? First, this leaves us with the insight that real-world indicators do not behave as neatly as theoretically derived concepts. However, this is only one part of the 9

10 story, since the deviances from the sum of 1 can be empirically also other than just marginal or meaningless. What is more is the insight that the ideal types are no true opposites to one another. To be sure, Schmitter and Streeck do not claim this anywhere in their text. However, this insight nevertheless has important consequences for Schmitter s and Streeck s graph: if the ideal types are not true opposites to one another (i.e., do not belong to the same fuzzy scale), then the logics are not true opposites to one another, either. If the logics were true opposites, then the combination of their extremes would again have to produce true opposites. Since the latter is not the case, the former cannot be said, either. But perhaps such a clear-cut statement was not even on Schmitter s and Streeck s agenda, although their reference to the Janus-like nature of [ ] associations in their role as intermediaries between at least two independently constituted, resourceful and strategically active sets of actors (Schmitter & Streeck 1999 [1981]: 19) suggests something of the kind. However, when Schmitter and Streeck established the labels of the logics, they did not even choose a terminology which reflects the oppositional character of the logics. An opposition between influence and membership or between goal formation and effective implementation is no semantic necessity. Also, from a fuzzy set perspective, the fuzzy negation of the presence of a strong logic of influence would be the absence of a strong logic of influence, which is not necessarily conceptually equal to the presence of a strong logic of membership. Thus, as a preliminary conclusion we can hold that the opposite logics are used in the literature in complementary ways, but that not even Schmitter s and Streeck s wording confirms this exclusive use of them as complementary categories. Being equipped with this insight, we can make use of the not-any-more orthogonal structure of the graph, adding some more insights from fuzzy set theory. This concerns, first of all, how to use these insights for statements about connecting the four logics with the four ideal types, on the basis of established fuzzy values for membership in the four ideal types. It is clear that assessing the presence or absence of, e.g., the logic of membership must be a function of the fuzzy values of the interest group under analysis with regard to governments and with regard to movements (since these are the corresponding ideal types to the logic of membership). But how shall the two fuzzy values be aggregated? One possibility would be the arithmetic mean which has become kind of the default option in index building. However, this would mean that, if an interest group is similar to a movement, but not to a government, the weak performance as a government reduces the strength of the logic of membership. However, such a conclusion does not make sense. The strength of the logic of membership in that case is due to the interest group being so similar to a movement. In other words, one of the two types is sufficient to be there in order to indicate the presence of a given logic. This idea of sufficiency is not only one of the cornerstones of set-theoretic social science analysis (Schneider & Wagemann 2012: 57), but is also being used in set-theoretic reasoning about concepts (Goertz 2005: 35ff.). 10 In set theory, these sufficiency relations between single sets are illustrated as unions 10 In order to understand better the concept of sufficiency in concept formation, the term of functional equivalence (Goertz 2005: 15) can be helpful. In our case, the two ideal types are functionally equivalent 10

11 of sets and are verbally connected through the logical OR (Schneider & Wagemann 2012: 45). The logical OR then leads to the so-called maximum rule which means that the membership of a case in a union of sets is the maximum of the individual memberships of the case in the constituting sets (Schneider & Wagemann 2012: 46). Applying this principle, the fuzzy membership values of associations in the ideal types can be used to derive via the maximum of the corresponding ideal types fuzzy membership values for the four logics, without having to recur to indicators for the logics. In the end, every association is characterized by four fuzzy values, one for each logic. We can now take Schmitter s and Streeck s original typology and develop profiles of single interest groups, by simply graphically connecting the various points which are identified through their fuzzy set memberships on the four logics. This will lead to four-corners geographical figures (which are only truly rectangular in very special settings). Figure 2: Examples of interest group profiles (1) These profiles similar to spider (or radar) charts in quantitative analysis can graphically describe the composition of associational logics. This is especially important, since the above has shown us that an association cannot only correspond to two logics (which would be the case if the pairs described opposites), but to up to all four (since the pairs do not describe opposites). At the same time, this mode of representing profiles also works for those associations which concentrate on just one logic. 11 There is yet another useful effect: it might happen that two (or more) given associations are characterized by profiles (derived in the described way) which are different with regard to the associational logic which they share. Another useful term to describe the idea is substitutability : with regard to the shared logic, both corresponding ideal types can substitute one another. 11 Of course, the profile can also capture those associations which do not concentrate on any of the logics. However, this only happens if the associations do not correspond to any of the ideal types. It is hard to imagine how such an interest group should empirically look like. 11

12 in size, but are, at the same time parallel (i.e., all four lines making up the profile are parallel to one another, and the profiles are stretches of one another), as in Figure 3: Figure 3: Examples of interest group profiles (2) In this case, interest groups have the same profile in kind, but are different with regard to their intensity in showing the profile. Similarities of associations have different intensities. In other words, these profiles as a final result of our considerations clearly mirror the underlying principle of fuzzy sets, namely the combination of differences in kind and in degree. Actually, they become a kind of fuzzy profiles. 6 Concluding Remarks and Future Research Our paper departs from the observation of the notable and path-breaking contribution which the neo-corporatist agenda of the 1970s and 1980s had on interest group research, both from a conceptual and a substantial point of view. First and foremost, we pointed to the project design of Schmitter and Streeck (1999 [1981]). More precisely, we referred to their development of two different sets of (supposed) opposite associational logics, namely the logic of influence vs. the logic of membership, and the logic of effective implementation vs. the logic of goal formation. We took up Schmitter s and Streeck s idea to combine these logics into ideal types which received the labels of other types of corporate actors, namely governments, firms, movements, and clubs. In a subsequent step, we showed with the help of indicators for each of the four ideal types that they are no true opposites to one another. Consequently, the pairs of logics are no true opposites to one another, either. Taking advantage of this, we proposed a maximum 12

13 approach for the description of associational logics and developed profiles which characterize individual associations. These profiles can show on which logics associations place emphasis; the logic of fuzzy values which also capture differences in degrees also offers the possibility to show which associations are similar in kind to which logics, while representing them at different degrees. Notwithstanding, this study represents a very preliminary analysis, and further research on a number of issues is required. First of all, a further refinement of the indicators for the four quadrants is required: our proposal is only a first step towards a goal which is both necessary as well as very demanding. Second, and only after a good deal of consensus will have formed around concepts and indicators, it will be both necessary and very useful to proceed with the empirical analysis of interest groups, mainly in comparative perspective. In conclusion, still much has to be done in order to fully conceptualize interest groups and their underlying associational logics. This paper has not yet discussed whether other (implicitly) dichotomous pairs of logics would have to complement the existing proposal. We are only opening a discussion, adding some insights from novel methodological approaches; in other words, this paper is mainly an hommage to a path-breaking and nearly unique typological scheme which in its basics has survived for more than 30 years by now. 13

14 References Baroni, L., B.J. Carroll, A.W. Chalmers, L.M. Muñoz Marquez & A. Rasmussen (2014). Defining and classifying interest groups, Interest Groups & Advocacy 3(2): Baumgartner, Frank R. & Adriana Bunea (2014). The state of the discipline: authorship, research designs, and citation patterns in studies of EU interest groups and lobbying, Journal of European Public Policy 21(10): Baumgartner F.R., J.M. Berry, M. Hojnacki, D.C. Kimball and B.L. Leech (2009). Lobbying and Policy Change: Who Wins, Who Loses, and Why. Chicago: The University of Chicago Press. Beyers, Jan, Rayner Eising & William Maloney (2008). Researching Interest Group Politics in Europe and Elsewhere: Much We Study, Little We Know?, West European Politics 31(6): Binderkrantz, Anne S. (2009). Membership Recruitment and Internal Democracy in Interest Groups: Do Group-Membership Relations Vary Between Group Types?, West European Politics 32(3): Binderkrantz, Anne S., P. Christiansen & H.H. Pedersen (2014). Interest Group Access to the Bureaucracy Parliament and the Media, Governance forthcoming. Doi: /gove Child, John, Arthur Loveridge, & Malcolm Warner (1973). Towards an Organizational Study of Trade Unions. Sociology 7: Crombez C. (2002), Information, Lobbying and the Legislative Process in the European Union, European Union Politics 3(1): Czada, Roland (1994). Konjunkturen des Korporatismus: Zur Geschichte eines Paradigmenwechsels in der Verbändeforschung. In Wolfgang Streeck (ed.), Staat und Verbände. Opladen: Westdeutscher Verlag, pp Dahl, Robert (1961). Who governs? Democracy and power in an American city, New Haven, London: Yale University Press. Eising, Rayner (2007), Institutional Context, Organizational Resources and Strategic Choices: Explaining Interest Group Access in the European Union, European Union Politics 8(3): Goertz, Gary (2005). Social Science Concepts. Princeton: Princeton University Press. Grant, Wyn (1989), Pressure Groups. Politics and Democracy in Britain. London: Philip Allan. Halpin, Darren (2006). The Participatory and Democratic Potential and Practice of Interest Groups: Between Solidarity and Representation, Public Administration 84(4): Jordan, Grant & J. Greenan (2012). The Changing Contours of British Representation, in Halpin, D. & G. Jordan (eds), The Scale of Interest Organization in Democratic Politics: Data and Research Methods. New York: Palgrave Macmillan. Lazarsfeld, P.F. (1937), Some Remarks on Typological Procedures in Social Research, Zeitschrift für Sozialforschung 6: Maloney W., J. Jordan J. & A.M. McLaughlin (1994). Interest Groups and Public Policy: The Insider/Outsider Model Revisited. Journal of Public Policy 14(1): Moe, Terry M. (1980). The Organization of Interests. Incentives and the Internal Dynamics of Political Interest Groups. Chicago: The University of Chicago Press. Olson, Mancur (1971 [1965]). The Logic of Collective Action. Cambridge (Mass.): Harvard University Press. Ragin, Charles C. (1987). The Comparative Method. Berkeley: University of California Press. Ragin, Charles C. (2000). Fuzzy-Set Social Science. Chicago: The University of Chicago Press. Ragin, Charles C. (2008). Redesigning Social Inquiry: Fuzzy Sets and Beyond. Chicago: The University of Chicago Press. Salisbury, Robert H. (1984). Interest representation: The dominance of institutions, American Political Science Review 78(1):

15 Schmitter, Philippe C. (1974). Still the Century of Corporatism? The Review of Politics, 36, 1: Schmitter, Philippe C. (1989). Corporatism is Dead! Long Live Corporatism! Reflections on Andrew Shonfield s Modern Capitalism. Government and Opposition, 24: Schmitter, Philippe C. & Jürgen R. Grote (1997). The Corporatist Sisyphus: Past, Present and Future. EUI Working Papers, SPS 1997/4. Schmitter, Philippe C. & Wolfgang Streeck (1999 [1981]). The Organization of Business Interests: Studying the Associative Action of Business in Advanced Industrial Societies. Max Planck Institute for the Study of Societies. Discussion Paper 99/1. Schneider, Carsten Q. & Claudius Wagemann (2012). Set-Theoretic Methods for the Social Sciences. Cambridge: Cambridge University Press. Streeck, Wolfgang & Philippe C. Schmitter (1985). Community, Market, State and Associations? The Prospective Contribution of Interest Governance to Social Order. In Wolfgang Streeck & Philippe C. Schmitter (eds.), Private Interest Governments. London et al.: Sage, pp Truman, David (1951). The Governmental Process. New York: Knopf. 15