The Representative Potential of Interest Groups Internal Voice in Post‐Communist and Western European Countries

Why do some interest group systems provide group members with more elaborate voice opportunities than other systems? We argue that evaluating membership voice is important for understanding the representative potential of interest group systems. An adequate understanding of “voice” forms the basis of “context”‐embedded assessments of benchmarks such as interest group bias, interest group representational distortion, and interest group‐driven policy overload. We examine two competing hypotheses on the differences in internal voice in Eastern and Western Europe. Primarily, case‐specific arguments lead us to expect a weaker internal voice in post‐communist Eastern Europe compared to Western Europe. Conversely, some theoretical approaches, such as population ecological organisational theory, lead us to expect a relatively weak membership voice in the organisationally saturated Western European systems. We assess these two hypotheses on the basis of an international survey of interest group leaders and observe, in line with the population ecological hypoth‐ esis, that members of Western European interest groups, compared to those in post‐communist countries, are perceived as having less influential voices in internal decisions on policy positions. We conclude, neither optimistically nor pessimisti‐ cally, that there is a meaningful representative potential of interest group systems supporting democratic societies, also, or even especially, in the post‐communist countries studied.


Introduction
One core question in political science concerns the relation between organised interest representation and democratic governance.In the early 1950s, American scholars aimed to explain democratic performance in order to ultimately advise Western governments on their foreign policy in relation to non-democratic countries (Almond, 1958).In the view of Almond and other scholars, the intermediate position of interest groups between public opinion, parties, and government makes interest groups crucial for "a more complete and systematic conception of the political process as a whole" (Almond, 1958, p. 271).The central idea is that a conceptual difference exists between countries on the basis of group politics rather than "only" in terms of formal governmental procedures.By implication, the quality of group politics positively shapes the quality of democratic politics.However, this central attention to groups contrasts sharply with some contemporary views on politics.For instance, the core challenge (or solution) to democracies, currently labelled "democratic backsliding," is firmly identified as being outside of "group politics" but clearly in the area of party politics ("populism") or executive politics (authoritarian leadership; e.g., Waldner & Lust, 2018).Interest group politics is seen as secondary to other systemic characteristics or are viewed as a problem (Mounk, 2018), not a solution, for democratic governance.
In this article, we extend earlier studies (e.g., Berkhout, Hanegraaff, & Maloney 2021;Binderkrantz, 2009;Bolleyer & Correa, 2022;Stavenes & Ivanovska Hadjievska, in press), further develop a so-called organisational view on interest group politics and situate interest groups more centrally in an understanding of the quality of democracy.This contribution of the study is especially relevant due to the process of democratic backsliding that is currently occurring across several EU countries.Interest groups are notoriously absent in debates on these developments.For instance, prominent "democracy" indices, including the Freedom House (2022) Index and the index of the V-Dem project (see Section 3.6.0.5 of Coppedge et al., 2021), only include important (e.g., Bolleyer, 2021) but relatively superficial and low-impact measurements on the regulation of civil society organisations and do not consider professional or business interest associations.From one perspective, interest groups may actively try to challenge democratic backsliding by mobilising citizens and public opinion.In this way, these groups may counteract the centralisation of power within states and strengthen the societal control of the state (e.g., Acemoglu & Robinson, 2019).Conversely, interest groups may increasingly come under subtle executive control and serve as top-down extensions of one or a select number of powerful political leaders.In this case, the groups become an extension of the state, which may be increasingly single elite-dominated, and they may not act as a societal constraint on state power to the same extent.In the latter scenario, membership participation decreases in importance, and this may reinforce processes of democratic backsliding.The degree to which interest groups play these roles in different contemporary democracies remains unclear.
This article provides a first and partial answer regarding some issues related to this question.More precisely, we analyse the representative potential of interest groups across Western and Eastern European countries by investigating the extent to which these groups provide their members with the opportunity to voice their policy preferences.Our starting point is that democratic governance may profit from a vibrant interest group system.Following Hirschman's (1970) notions of membership exit and voice, such vibrant systems combine two important characteristics: at the macro-level, competition among groups with different political views (hence, plurality and diversity), and, at the meso-level, the internal voice of members in organisational decisionmaking (hence, participation).A diverse system provides free choice to the potential members to join a group of their liking, and an internal voice in organisational decision-making ensures that the groups remain sufficiently independent from state actors.A lack of external competition and internal voice make interest group sys-tems vulnerable, whereas the presence of these two features strengthens the interest group systems as representatives of society.
Former research has shown that the Western European interest group systems are denser and more diverse compared to those in Eastern Europe (Hanegraaff et al., 2020), which suggests that citizens in the West have more choices for joining groups.However, the extent to which membership involvement varies across Western and Eastern countries is much less clear.Hence, we seek to explain the varying patterns of involvement between countries.Our explanatory model juxtaposes two plausible research outcomes.Firstly, we discuss the argument that interest groups in post-communist systems have a relatively weakly developed responsive internal organisational culture due to the relatively limited interest on the part of (potential) members.Secondly, in contrast, population ecology approaches predict that post-communist European groups may be more responsive to their members because these organisational systems have had a shorter time to develop a wide variety of organisational forms with varying degrees of membership involvement.Our empirical analysis relies on Comparative Interest Group Survey data from eight European countries.In this article, we first elaborate on the relation between interest group systems and democratic governance.Subsequently, we discuss the two main hypotheses, after which we test them using several multivariate analyses.We find that organisational representatives in post-communist countries, most notably Poland and Lithuania, more frequently identify members as influential compared to the organisational representatives in the other studied countries.This contrasts with accounts indicating the weak nature of civil society in post-communist Europe.We end by presenting several concluding remarks and certain avenues for future research.

What Does a "Democratic" Interest Group System Look Like?
The contemporary research on interest groups is rooted in the notion that the quality of interest representation largely relies upon and can be explained by its relationship to several institutional, issue-specific, or broader "contexts" (e.g., Klüver et al., 2015;Lowery & Gray, 2004).In order to qualify as "democratic," interest group systems should be "unbiased" in relation to salient interests in society (e.g., Lowery et al., 2015), avoid encouraging divisive political party polarisation (e.g., Berkhout, Hanegraaff, & Statsch, 2021), and be congruent with large majorities of the public (e.g., Rasmussen, 2019).These contingent and contextual implications are assumed to vary across different stages of the so-called influence production process (mobilisation, population, strategies, influence).Methodologically, this variation creates a plethora of potentially relevant benchmarks for assessing the democratic role of organised interests.
In addition, these "contextual" benchmarks force scholars to explore or assume the potential outcomes of group politics relatively far down the causal chain-for instance, Olson's (1982) "sclerosis" hypothesis.
We think that it is more productive to identify a benchmark that is proximate to interest groups rather than a characteristic of politics or society more broadly.Therefore, we aim to conceptually develop and empirically assess the characteristics of interest groups themselves.Instead of analysing substantive benchmarks, such as the diversity of the group system in terms of policy views and preferences, we explore the organisational mechanisms.More precisely, we investigate how interest groups involve their members in internal decision-making processes, which is referred to as "internal voice."Before outlining our focus on "internal voice" in greater detail, we first discuss some context-oriented norms of democratic group politics.We cluster these norms into three categories: biased representation, distorted representation, and policy output-centred representation.Although this discussion may seem somewhat removed from our main objective, which is to explain why some interest groups provide more internal voice than others, we discuss these contextual benchmarks at length to illustrate that, ultimately, they all rely upon the conceptual and empirical existence of "a transmission mechanism" of interests, for which membership involvement is critical.
Firstly, the most commonly discussed norm relates to bias.Interest group scholars routinely rely on variants of Schattschneider's identification of "bias" in interest group politics and his challenge of the earlier pluralist idea that the interest groups system should more or less reflect the salient interests in society (see discussions in Lowery et al., 2015).Such bias may arise at several stages; for example, citizens may not be fully aware of their interests, latent interests may not be organisationally articulated, citizens may be out-voiced by business interests, policymakers may be selective in their hearing, and status-quo policy programmes (and their supporters) may be resistant to change.All these mechanisms reduce the likelihood of societal concerns being adequately heard in politics or transmitted by the group system.Instead of reflecting societal interests, the group system may reinforce the unequal distribution of benefits that may arise from public policy programmes.Hence, "bias" may provide information about the quality of interest representation (e.g., Lowery et al., 2015).However, it is also a demanding benchmark given that it always requires the identification of a meaningful connection between the issues publicly voiced in group politics and citizen preferences.Establishing this connection empirically is not an easy task as it requires the operationalisation of the connection between public opinion data and interest group positions (e.g., De Bruycker & Rasmussen, 2021;Flöthe & Rasmussen, 2019).Therefore, often, these connections are based on theoretical assumptions, such as qualifying "business bias" as normatively prob-lematic.In situations of "democratic backsliding," political bias in interest group systems may be assumed to arise from the strategic initiatives of officeholders that aim to reduce the policy access and voice of political opponents by means of restrictive regulation or other state action.
Secondly, Klüver (2020, p. 980) noted that "if interest groups manage to influence party policy at the expense of voters, democratic representation is seriously undermined by lobbying."Therefore, interest groups distort the representational activities of other organisations, especially political parties, as well as other intermediary institutions, such as the media (e.g., Trevor Thrall, 2006) or consultation venues established by governments (e.g., Arras & Beyers, 2020;Binderkrantz et al., 2021;Fraussen et al., 2020).Indeed, political representatives may receive "signals" from interest groups and mistakenly interpret these signals as support from public opinion (e.g., Rasmussen & Reher, 2019).Instead of representing their "true" constituency (for instance, voters), policymakers represent group positions.Hence, elected politicians are not political representatives but policymakers acting on behalf of some specialised interest group.As with bias, this benchmark also conceives the democratic quality of the interest group system as being external to interest groups themselves because it must be seen in relation to the responsiveness of political parties or other representative institutions.This contextualised benchmark of "distortion" is also empirically and conceptually demanding to analyse, especially in cross-country comparisons.
Thirdly, interest groups may be judged by the plausible outcomes of their engagement in the policy process.In broad terms, scholars have noted that interest groups are instrumental for the realisation of broadly agreed public policy objectives, such as economic growth or low unemployment; however, they may also make the policy process inefficient and inflexible to changing circumstances (e.g., Anderson, 1977;Olson, 1982;Schmitter, 1977Schmitter, , 1981)).According to Anderson (1977, p. 148), "interest representation is legitimate only insofar as it is instrumental to the achievement of stipulated public objectives."Olson (1982) assumed that "nations" want to "rise" economically and suggested that any group activity, according to his "institutional sclerosis hypothesis," is unlikely to be instrumental in that regard as it would interrupt the efficient allocation of (public) resources.Similarly, Schmitter (1977Schmitter ( , 1981)), took economically efficient public policy as a meaningful benchmark.In his view, interest groups, particularly when they are "encompassingly" organised into associations, potentially create opportunities for the effective management of economies.This debate has resurfaced in recent studies on stakeholder engagement in, for instance, regulatory consultations (e.g., Fraussen et al., 2021).As with distortion and bias, interest groups are primarily judged on the basis of consequences, such as policy outputs or, in contemporary terms, regulatory legitimacy (e.g., Braun & Busuioc, 2020), that are external to the interest organisation itself.
These three contextually based benchmarks of interest representation (bias, distortion, and policy output) depart from the idea that interest groups ultimately connect important societal interests with public policy.We think that an alternative assessment of the representative quality of group politics should focus on organisational qualities that facilitate such an intermediary function.Such a perspective can be conceptualised independently from the plausible causes and implications of interest group politics.In simple terms, interest groups should be able to respond to changes in the relationship with their members, supporters or, more broadly, their constituents.Hirschman's (1970) Exit, Voice and Loyalty presents a classic conceptualisation of organisational responsiveness that can be used in interest group studies (e.g., Barakso & Schaffner, 2008;Berkhout, Hanegraaff, & Maloney, 2021).Hirschman noted that consumers of a good respond to dissatisfaction by choosing alternative suppliers and exiting or discontinuing a prevailing relationship (see also Warren, 2011).For instance, Richardson (1995) noted that in the "market for political activism" in the 1970s and 1980s, large numbers of citizens quit the presumably overly traditional, formal, and materialistic political parties in favour of social movements focussing on "new" issues and offering more informal, low-threshold participatory options.Similar competition for members occurs among interest groups.Organisational leaders are incentivised to prevent exit and invest in means to improve the long-term engagement of members, such as through membership magazines, outreach, and opinion research, often combined with continuous membership recruitment campaigns (Jordan & Maloney, 1998).This exit mechanism works at the level of organisational communities of "like-organisations.''For instance, environmentally concerned citizens may move from supporting one environmental NGO to supporting another environmental NGO (but not to support a business association).Organisational leaders can respond more effectively if they address the potential threat of members exiting their organisation in favour of a similar one.We concur with these arguments on exit, but as this hypothesis has been already broadly tested in the literature (for an overview, see Berkhout, Hanegraaff, & Maloney, 2021), we further develop the second component voice of Hirschman's model in this article, namely "voice." We conceptually and empirically focus on associations with firm or citizen members and on organisations with supporters that are (potentially) politically active (Baroni et al., 2014;Beyers et al., 2008).This research concerns organisations that claim to be representative of their (potential) members, as reflected in some organisational characteristics and procedures (Jordan et al., 2004;Warren, 2001).We also include organisations that derive their legitimacy largely from the cause they represent (e.g., Montanaro, 2017;Nuske, 2022), which are sometimes labelled "solidarity organisations" or "cause groups" (Halpin, 2006), even though the internal voice provided may have a somewhat different function in these organisations: As observed by Berkhout, Hanegraaff, and Maloney (2021), these types of internal voice are not fundamentally incomparable in empirical terms.We exclude a plethora of organisations (such as firms and semi-public agencies) that are active in the policy process and act as pressure participants (Jordan et al., 2004) but have internal processes that are not directly comparable with the membership voice practices within associations.

Internal Voice as an Indicator of Representative Norms Within Interest Groups
Hirschman's understanding of voice implies that when the consumers of a good, in our case members of interest groups, are not fully satisfied, they do not choose to quit the association but may voice their views internally.For example, when a general practitioner and a member of a professional association of doctors experience a lack of qualified interns, they may become active in a relevant sub-committee within the professional association and make the internship issue a higher priority of the association.As noted by Warren (2003, p. 48), "the associational way of organizing common purposes is inherently legitimate, since people choose their collective projects and willingly engage with others," and, again according to Warren, associations outperform faceless markets and representation-based forms of political deliberation, since "deliberative elements of a democracy can only be organised along associational lines." We are not alone in our focus on voice practices as an important benchmark for the representative quality of interest groups (e.g., Albareda, 2018;Bolleyer & Correa, 2022;Fraussen et al., 2021;Heylen et al., 2020;Warren, 2001).Voice is understood to have both formal and behavioural components.Formal voice refers to the organisational rules on the control that the membership has over important decisions (board appointments, strategic policy decisions, etc.), and behavioural components include the extent to which substantial parts of the membership are actively involved in decision-making, including at lower levels, such as in local branches or topic-specific committees.
Several studies have identified professionalisation (e.g., Bolleyer & Correa, 2022;Heylen et al., 2020) and political accommodation as threats to (e.g., Schmitter & Streeck, 1999) or incentives for (e.g., Grömping & Halpin, 2019) membership involvement.Business and professional associations also seem to offer more voice to members than citizen groups do (Berkhout, Hanegraaff, & Maloney, 2021), and the multi-layered nature of EU politics creates particular challenges and opportunities regarding membership involvement (Albareda Sanz, 2021;Hollman, 2018;Ronit, 2018).Membership involvement increases the congruency between interest group public policy positions and public opinion (Willems & De Bruycker, 2019).Voice, as conceptually noted by Hirschman, heavily depends on exit options.When voting with one's feet by moving to a different group is not possible, such as if there are few alternative groups to join, one is forced into using voice as the only active means of response (Warren, 2011).In organisational communities with effective alternatives, it is, thus, more common to find organisations with only limited internal voice options (Berkhout, Hanegraaff, & Maloney, 2021).In this article, we use organisational voice at the aggregate level of national interest group communities in order to assess the nature of interest group politics and its potential contribution to democratic governance.
In short, internal voice opportunities increase the responsiveness of interest organisations to critical membership environments.More responsive interest organisational communities in which the average group shows high membership involvement improve the intermediary function of interest groups in democracies.In this case, the intermediary structure of the interest group system may be strengthened, thus making it less dependent on the state and more society-driven to such an extent that it might counter tendencies towards democratic backsliding.Many decades ago, Truman (1959, p. 491) identified the role of an "intermediate structure" as an indispensable part of democracy and noted that organisational leaders are responsible for reducing the vulnerability of the democratic system to "demagogic leaders whose actions may constitute a threat to the system of procedures."Responsible association leaders require support from their members; specifically, when the leadership is heavily incentivised to follow the opinions and preferences of the organisational members it is less likely to become a spokesperson of some political elite.Concurrently, groups are also places of power politics where, as a "voluntary" exchange, if the members are heard and are involved in developing a group position, they may consent to follow the leaders.According to Streeck and Kenworthy (2005), groups behave in a disciplinary manner, meaning they exchange meaningful voice for control over important decisions.Indeed, interest organisational systems with effective internal voice mechanisms are more likely to engage in democracy-defending roles, thus, in the long-term, guaranteeing some meaningful diffusion of power away from state actors (e.g., Acemoglu & Robinson, 2019).Hence, internal voice opportunities are important for stabilising democratic systems and preventing democracies from backsliding into authoritarian modes of governance.The descriptive differences between countries in the degree of membership voice are, thus, valuable in and of themselves.Given that there is currently a lack of systematic comparative interest group studies, there is also relatively limited system-level theory regarding country differences with respect to internal voice.Moreover, the "classic" typologies of pluralism and cor-poratism are too encompassing to derive observable implications from, do not easily match the wide range of contemporary European countries, and, in countries commonly identified as typically pluralist or corporatist, also seem decreasingly valid in answering research questions on the structures of organisations or policy access (e.g., Aizenberg & Hanegraaff, 2020).
Why do some interest group systems provide more elaborate voice opportunities to members compared to other systems?In light of the limited system-level theory, we depart from the idea that the fall of communism in 1989 critically affected the interest group systems in Eastern Europe but affected those in the rest of Europe to a lesser extent, if at all.We identify case-specific circumstances that partially arose from the particularities of the pre-1989 systems present in Eastern Europe.We label these countries "post-communist" and "Western" systems in line with the main division within Europe during the Cold War era.We distinguish between "qualitative" implications, related to the characteristics of the pre-existing regimes and the nature of the transition, and "quantitative" implications resulting from the actual time passed since the transition.
Firstly, scholars have identified important qualitative characteristics of the 1989 political revolutions that may have important consequences for organised interests: (a) relatively low levels of political participation among citizens and (b) relatively "unsophisticated" internal voice structures due to the relatively limited dependence on membership fees.To start, as noted by Howard (2011, p. 134; see also Howard, 2003), "after the 'revolutionary' moment had passed, people left the streets and their civic organizations, leaving their societies largely passive and depoliticized."In the Eastern European waves of the World Value Survey between the early 1990s and the following decades, citizens indicated relatively low levels of membership, participation, and trust in several types of social, political, and civil associations.Comparative protest event data indicate relatively weak development of social movement protest activity independent from party politics (Borbáth & Hutter, 2021).These patterns seem to be a recurring finding, though there is scholarly discussion regarding the exact magnitude, causes, and consequences of these patterns of low participation (e.g., Ekiert & Kubik, 2014;Ekiert et al., 2017;Meyer et al., 2020;Navrátil & Kluknavská, 2020).
Related to our study of internal voice, Novak and Komar (2020, p. 650) noted that interest groups contribute positively to democratic governance when the "members are actively included in the internal decisions of interest groups" (in addition to the inclusion of groups in policy processes, which is another important benchmark of democratic governance).Comparing Montenegro and Slovenia, they concluded that "Montenegrin interest groups have been a tool of influence and democratisation primarily on behalf of the international community, [and subsequently] their internal democracy is less sophisticated than is the case in Slovenia" (Novak & Komar, 2020, p. 650).The transitional status of post-communist countries led to the presence of European and international subsidies that were commonly intended to support the democratic transition process because interest groups were seen as vehicles for democratic governance.These subsidies may have plausibly affected the internal processes of associations to a greater extent than the (also substantial) "domestic" subsidy programmes in Western European countries (also note the complex relationship with professionalisation; see, e.g., Dobbins et al., 2022;Europeanisation, Cekik, 2017).However, the greater reliance on external donors has made many groups less dependent on their membership.The aspiration to receive financial support from and be responsive to European and international donors among important parts of interest group communities in post-communist countries may potentially reduce the internal voice of the members.These qualitative implications of the 1989 transitions in terms of high external organisational dependency (donors) and potentially passive membership attitudes lead to the following hypothesis: H1: Interest groups in post-communist countries offer fewer internal voice opportunities to their members than interest groups in Western Europe.
A second hypothesis follows from the fact that the 1989 start of the transition period means that, quantitatively, there has been relatively little time for the maturation of the organisational system in post-communist systems compared to Western countries with a longer time period of systemic stability.We use arguments from population ecological studies about the effect of time on changes in organisational systems.To start, in earlier studies, interest group system "maturation" was identified as a mechanism that explains differences between group systems in Eastern and Western Europe (e.g., Hanegraaff et al., 2020).Specifically, organisational systems require time to develop, both regarding the number of organisations up to a saturation point and the organisational diversity in terms of filling particular organisational niches (e.g., Aldrich et al., 1994;Gray & Lowery, 1996).We assume that the start date of interest organisational systems largely mirrors the politicalinstitutional upheaval of 1989.This assumption is similar to Olson's (1982) choice to use the end of the Second World War as the starting date for investigating the German interest group system (see also Labanino et al., 2021;Unger & van Waarden, 1999).This choice means that the Central and Eastern European systems have had around 30 years to develop, whereas the systems of the other countries studied have had around 70 years (although the "age" of the Portuguese system in our sample falls somewhat outside this pattern).
How does system maturation affect internal voice?Increasing maturation is related to competition and associated specialisation, and this commonly increases the variety of organisational formats (Aldrich et al., 1994).This variety of organisational formats increases the likelihood of relatively democratic associations, such as organisations with strong membership involvement, being outcompeted by less democratic ones.This competition in terms of membership voice arises from variations in the wishes of potential members who may sometimes seek expressive benefits and internal voice and, at other times, may be satisfied with donating only.As argued by Hirschman (1982), citizens shift their participatory preferences; under some circumstances, at certain moments of a "cycle," they may wish to voice their views, whereas, at other times, they may not feel the need to be involved in the associations of which they are members.These fluctuations in dissatisfaction within individuals and, plausibly, between individuals produce a fertile ground for important diversity in organisational forms, including non-collective action, top-down structured associations and "flat" personal network-like organisational formats.
An important source of organisational diversity arises when cause groups are established that seek citizen donations rather than voting membership fees to offer low-threshold engagement opportunities to citizens, as well as when a professionally run "non-membership advocacy organisation" enters an interest group population (e.g., Minkoff et al., 2008;Walker et al., 2011).Similar dynamics may occur in relation to business interest representation.To illustrate, at the height of a cycle of attention, firms, while keeping their membership to a specialised association, may establish a public affairs department and start lobbying outside of business interest associations.Subsequently, such diversification practices may trigger a wider variety of organisational forms in a given population (e.g., Aizenberg, 2021;Salisbury, 1984).Increasing variation takes time, as "cycles" of societal niche formation occur only sporadically, and organisational leaders develop organisational forms on a trial-and-error basis.The long-term nature of the trend towards increasing diversity in organisational form leads us to expect that interest organisations from postcommunist countries may be more responsive to their members.Indeed, in post-communist countries, there has been less time for less democratically organised interest groups, such as professionalised associations driven by experts, to embed themselves in the organisational populations to the same extent as has happened in Western Europe, which leads to an alternative to our previous hypothesis: H2: Interest groups in post-communist countries offer more internal voice opportunities to their members than interest groups in Western Europe.

Design
We used data from the Comparative Interest Group Survey (2020), which comprises evidence collected from surveying Dutch, Belgian, Portuguese, Swedish, Polish, Slovenian, and Lithuanian interest groups (Beyers et al., 2020).The country surveys were in the field for a couple of months in the time period between 2016 and 2018.The respondents had leading positions in national membership associations of individuals or companies with potential interests in public policy.The respondents were asked about organisational characteristics, political strategies, and their relations with political parties, parliament, government ministries, and agencies.The survey questionnaire was translated from an English language EU-oriented version by each of the research teams and adapted to the respective national contexts.Overall, the response rate to the survey was 36%, which is relatively high compared to other online surveys in this field (Marchetti, 2015).
The concepts introduced earlier were operationalised as follows (see Table 1 for summary statistics of the indicators used).The dependent variable "internal voice" was operationalised as the ability of members to "influence the policy positions of interest groups."We focused on how organisational leaders perceive the impact of members on the policy positions of the organisation with regard to the political agenda rather than in relation to "internal" issues.The internal consensus formation among members on policy positions potentially may involve several subsections of the association and may affect the success of interest groups in policy circles (in terms of access see Grömping & Halpin, 2019; in terms of policy influence see Truijens & Hanegraaff, 2021).Specifically, we asked interest group leaders the following question: Thinking about your organisation's position on public policies, how would you rate the influence of your membership?The respondents could indicate that the members were very influential, somewhat influential, not very influential, or not at all influential.We considered social desirability bias among the particular respondents, with some leaders potentially emphasising their own vision and control (and, thus, underestimating members' influence) and others potentially wishing to highlight their democratic credentials.Indeed, some group leaders may, because of a normative bias, overstate the membership influence, whereas others may underestimate the membership influence.We are reasonably confident that such social desirability is unlikely to be very problematic, as our indicator strongly correlated with a number of other questions, such as those related to the influence of members on the strategies of the organisations (r = 0.61, p = 0.000) or membership elections for the executive boards.Precisely, stronger formal opportunities for members were associated with higher perceived membership influence.This correlation supports the validity of our measurement for assessing the formal (opportunities for influence), behavioural (actual use of these opportunities), and anticipated (leadership expectations regarding the possible views of members) components of "internal voice."The correlation also increases our confidence that our operationalisation is comparable to that of earlier studies, such as those based on executive board elections (Stavenes & Ivanovska Hadjievska, in press), several formal organisational features (e.g., Albareda, 2018Albareda, , p. 1218)), and some combination of statutory formal membership influence indicators and leadership perceptions regarding the involvement of members (e.g., Binderkrantz, 2009, p. 669;Bolleyer & Correa, 2022).
Our main independent variable was the geographical region in Europe: a post-communist or a Western country.For Western countries, we clustered the responses of the following countries: the Netherlands, Belgium, Portugal, and Sweden.For post-communist countries, we grouped the responses of the Lithuanian, Slovenian, and Polish participants.We had more participants from Western European countries compared to post-communist European countries, with 1,710 and 707 participants, respectively.We controlled for group type, professionalisation, resources, identifying as a lobby or service provision organisation, the access of the organisations, the level of competition experienced by the organisations, and whether organisations were active in social policy areas or not.As identified in the studies noted earlier and below, each of these variables could explain the variation in membership voice.Firstly, the group type variable was coded based on the organisational websites by the researchers from the respective national teams.The coding scheme included eight categories (in parentheses are the number of observations for the all-countries sample): business groups (n = 549), professional associations (n = 512), labour unions (n = 119), identity groups (n = 296), cause groups (n = 542), leisure groups (n = of institutions (n = 79), and "other" (n = 59).Berkhout, Hanegraaff, and Maloney (2021) found that business organisations give more voice to their members compared to citizen groups, so we controlled for this.Secondly, we controlled for the level of professionalisation by considering the way the organisation made decisions, the criteria used when hiring staff, the staff training, and whether the employees were inclined to pursue a career within the organisation.More professionalised organisations may prioritise managerial decision-making over "inefficient" membership co-decisions (e.g., Bolleyer & Correa, 2022;Heylen et al., 2020).For resources, we utilised a question identifying the number of paid staff in the organisation.We conducted a log transformation on this variable due to the skewed nature of the responses.Larger organisations may be less responsive to individual members because of bureaucratisation, which is referred to as Michels' "iron law of oligarchy," among other reasons (e.g., Rucht, 1999).We also controlled for organisations identifying as lobby organisations or as being more service driven.This distinction was based on a question in the survey that asked the organisations to identify whether they were involved in a set of activities, including lobbying, research, promoting volunteering, and many more.If organisations indicated that they do not lobby, we treated this as an organisation whose main aim is to deliver a service to its members.There are inconclusive theoretical arguments on this distinction, but Bolleyer and Correa (2022) found that organisations that are more service-driven are less likely to be responsive and open to members.The fifth control variable was the access of the organisations to the policymaking process.We utilised a question examining the frequency of interest groups' access to policymakers (being invited by policymakers to provide policy input), including having contact with policymakers on a weekly, monthly, quarterly, or yearly basis, or no contact with policymakers.We included this variable as insiders are much closer to the policy process, might be more vulnerable to "co-optation" by policymakers, and, thus, become less responsive to the members (Bolleyer & Correa, 2022).Conversely, membership involvement in these organisations may provide such groups with additional political leverage and, thus, gives leaders an incentive to organise "voice" practice (e.g., Grömping & Halpin, 2019).The sixth control variable was the amount of competition faced by the groups to acquire resources.As discussed in the theoretical section, organisations that are in more competitive environments are less likely to provide a voice to their members.As the density in a system increases, organisations have to specialise in lobbying to achieve better results, which leads to less voice for the members (Berkhout, Hanegraaff, & Maloney, 2021).Finally, organisational features vary substantially between policy fields (e.g., Berkhout et al., 2017).We distinguished between social and economic policy fields, expecting the existence of closer relationships with members in the social field compared to the economic policy field.This variable was based on a question that asked respondents to indicate the policy fields in which the organisations are active.The first group (social policy) included health policy, gender policy, social policy, consumer protection, citizen's rights, and human rights.The second group (economic policy) included economic policy, fiscal and monetary policies, energy policy, foreign policy, defence policy, transport policy, and agricultural policy.

Results
In this section, we discuss our empirical findings.We utilised OLS regressions with robust standard errors.The results presented in Table 2 provide a clear demonstration of the factors that are important for internal voice.As found by other researchers (e.g., Berkhout, Hanegraaff, & Maloney, 2021), internal voice is more apparent among business organisations compared to citizen groups, such as identity and public interest groups.Moreover, resources are an important factor for voice.When groups become larger, they also become more detached from their members.The same pattern was identified for competition; specifically, as groups face more competition, they become more detached from their members, which is in line with Hirschman's argument on the relation between exit and voice, the neocorporatist argument on the logic of membership in the context of representational monopolies (e.g., Streeck & Kenworthy, 2005), and recent studies on internal voice (e.g., Berkhout, Hanegraaff, & Maloney, 2021).Similar to earlier studies, such as by Bolleyer and Correa (2022), we found that organisations that identify as advocacy/lobby organisations and those that are more frequent participants in policy (political insiders) are more likely to provide their members with a voice.Finally, organisations active in social policy fields are more responsive to their members compared to similar types of organisations active in economic fields.Next, we discuss the answer to our main question regarding whether organisations in Western countries are more responsive to their members compared to their counterparts in post-communist countries.We found that, contrary to our hypothesis (H1), organisations in Western countries are not more responsive to their members than those in Eastern countries.This is clear from the positive and significant coefficient reported in the top row of Table 2. Figure 1 indicates the strength of the effect by means of a plot of the predicted values.Indeed, we observed that interest groups in Western countries scored on average 3.08 for the amount of voice members have, while in Eastern countries, this score was 3.31.The difference corresponds to roughly a third of a standard variation in the answers respondents gave, which is not a very large difference but still represents a substantial effect.These findings support the populationecological hypothesis focusing on the maturation of the group system and suggest that the particular characteristics of the 1989 transition have not limited the participatory voice practices of group members.

Western Europe
Importantly, we assessed whether all countries in post-communist Europe differ from all countries in Western Europe.Specifically, we examined the argument that Slovenia's Cold War experience (as "socialist" rather than "communist") may have been different from the experience of the other two post-communist countries, which may make it more likely for Slovenia to have a more diverse organisational system with greater variation in voice options (Fink-Hafner, 2015;Novak & Fink-Hafner, 2019).In terms of our selection of Western European countries, we focused on Portugal, as its shorter time for organisational system maturation may have limited the variation in organisational format, with "standard" membership-controlled associations potentially being more dominant.To this end, we conducted a separate regression analysis (see Supplementary File) in which we used individual country dummies.Lithuanian organisations provide the most voice to their members of all organisations by a significant amount.Additionally, Polish organisations provide, on average, the second highest level of voice to members across all the studied countries (for an interesting interpretation see Pospieszna & Vetulani-Cęgiel, 2021).This result fits the population-ecological "maturation" theory and confirms our main analysis.However, the results, indeed, indicate that Slovenia is an outlier.Organisations in this country provide the fifth lowest level of voice to their members, with the level being close to the average level of voice in Western European countries.In particular, organisations in Slovenia provide more voice provided than those in Belgium and Portugal but less voice than those in Sweden and the Netherlands.The particularities of the organisational development in Portugal (with strong political party ties) seem to have produced organisations with limited voice (e.g., Lisi & Loureiro, 2022).Overall, this critical addition shows that researchers should pay attention to country differences within Eastern Europe as well.While the overall trends may support the argument provided in this article, there are some critical differences across countries.Therefore, new research should analyse these nuances and specificities.Indeed, this work confirms the relevance of the several studies in this thematic issue that looked into such variation.

Conclusion
We started this article by identifying the need to more explicitly understand the quality of group politics in assessments of the quality of democracy.We suggest that one most productive way to do so is by focusing on the organisational transmission qualities of interest group politics rather than, or at least prior to, aspiring to measure the plausible outcomes of interest group activities in terms of bias, distortion of representation, or policy overloading.We used Hirschman's classic argument of organisational voice and effective exit in order to conceptualise the key dimensions of transmission qualities in interest group politics.
Our research design focused on "voice" and compares post-communist and Western European countries.This comparison allowed us to investigate case-specific arguments on the historical persistence of practices that began in the time period surrounding the 1989 revolution, namely the communist legacy of "passivity" on the part of members in post-communist Europe and the organisational resource dependency on external donors rather than membership.By comparing post-communist countries with Western Europe, we also explored the effects of the different levels of system maturation, with comparably higher levels of interest group competition and specialisation present in Western Europe.
Our findings indicate that we should reject the core implications of the hypothesis that emphasises the particular qualitative characteristics of the 1989 transition.The findings suggest that, for interests organised into interest groups, members of interest groups in Eastern European countries have more internally organised influence on policy-related organisational decision-making compared to their Western counterparts.This is in contrast to existing empirical studies (e.g., Novak & Komar, 2020) and somewhat pessimistic case-specific accounts of the (limited) vibrancy of post-Communist interest group systems (e.g., Howard, 2011).Concurrently, the outcome supports the theoretical arguments on organisational maturation (Hanegraaff et al., 2021); in the plausibly more saturated systems in Western Europe, we observed a lower degree of internal voice and lower levels of membership influence on policy-related decision-making.
There is no simple answer to the question of whether our findings imply that the internally more representative and potentially more democratic Eastern European interest groups form an effective barrier against any tendency towards backsliding.Firstly, there are reasons to be optimistic.The interest group population in post-communist countries comprises a substantial number of groups that are largely controlled by active members.The substantial internal voice indicates a strong commitment to internal democracy within interest groups.This commitment should eventually strengthen the legitimacy of interest associations and may encourage the independence of interest groups in relation to the state and the dispersion of power away from the executive.
Conversely, and more pessimistically, our results also show that internal voice is weaker in the more competitive interest group systems in Western Europe.In this case, citizens may have a greater degree of choice among alternative interest groups.However, this more competitive system may reduce the strength of interest groups individually and collectively as a force for systemic stability and a counterbalance to executive power concentration and tendencies towards backsliding.Furthermore, our "static" comparative research does not allow us to evaluate the dynamics of backsliding (or the counter processes).Most importantly, we assume that the entry to the organisational communities studied through the establishment of associations or the development of political interests on the part of existing organisations is relatively open.However, in light of earlier studies on the shrinking civil space in some countries (e.g., Bolleyer, 2021), there may be a basis for pessimism in case this assumption of open entry does not apply to all countries.
It is clear that these processes deserve more scholarly attention and future study.For instance, it might be useful to understand more about the relative coverage of interest groups in different countries.For instance, in terms of professional associations, it would be useful to determine whether most professions are well-organised and represent the full breadth of a given profession.Another potential drawback of our analysis is that we focused mostly on group leaders.Indeed, it might be interesting to learn more the extent to which members practically make use of their (perceived) influence and "voice" their views within their association.With regard to "exit," further studies are needed to assess whether the anticipation of membership "exit" leads organisational leaders to be more responsive to (perceived) membership views.If that is the case, it would be possible to be relatively optimistic about competitive interest group communities, even in cases where voice mechanisms are limited.
Finally, previous interest group studies have broadly assessed "influence" in terms of preference attainment in relation to specific public policies.Much less research has been conducted at the system level and in relation to the overall functioning of democratic systems, meaning these topics require further study.For instance, when rioters stormed the US Capitol, lobbyists on the ground could do very little.However, interestingly, practically the whole Washington lobby community condemned the riot (e.g., National Institute for Lobbying and Ethics, 2021) and, to some extent, sought ways to support the democracy-saving elements in both parties, with several major corporations discontinuing the funding for individual Republican lawmakers who voted against the ratification of the election outcome.It could be studied how interest groups can support or undermine democratic institutions, especially because these responsibilities are something individual lobbyists are sufficiently aware of only infrequently.Indeed, as noted by Truman (1959, p. 489), the "holders of power" in the intermediate structure may consequently be "unaware of their positions' special vulnerabilities.Foremost among these is the possibility that the members of these elites will not see a threat to the system for what it is."

Figure 1 .
Figure 1.Predicted level of internal voice in interest groups in Western and Eastern countries.

Table 1 .
Descriptive statistics of the dependent and independent variables.

Table 2 .
Linear regression predicting the level of internal voice.