Exploring the domestic and international drivers of professionalization of Central and Eastern European interest groups

Abstract While there has been a veritable boom in literature on organized interests, their lobbying strategies, relationships with decision-makers, and their impact on policymaking, only a few studies have explored internal organizational developments and, specifically, the professionalization of interest groups. The present study focuses on the national and transnational factors driving the professionalization of interest groups in Central and Eastern Europe, a region previously neglected in much of the interest group literature. Based on a sample of more than 400 surveyed organizations operating in Poland, the Czech Republic, Hungary, and Slovenia in the healthcare, higher education, and energy sectors, we explore three bundles of factors potentially enhancing the professionalization of interest groups – organizational funding sources, national and transnational intergroup cooperation and organizations’ standing in the domestic interest group system. Our statistical analyses show that state subsidies and tight policy coordination with the state are crucial drivers of internal organizational professionalization, suggesting rather patronistic and symbiotic relationships between the state and certain organizations. However, our data also support the notion that interorganizational collaboration, both at the national and international levels, may also be key to organizational professionalization, enabling groups that lack close ties with the state to compensate their disadvantage with intensive domestic and international networking. The study is also among the first to link increasing professionalization with organizational population density.


Introduction
There is a longstanding consensus that civil society and interest organizations are weak and underdeveloped in Central and Eastern Europe (CEE) due to lasting legacies of civic oppression (Howard, 2003;Borragán, 2004;Ekiert and Foa, 2011). Under communism, associational life was essentially governed by the party state. Although these regimes promoted some civic organizations, for example sport and youth associations, participation was generally state-regulated and mandatory, while organizational leadership had little leeway to autonomously focus on organizational development.
Yet despite their historical weakness, civic organizations seeking confrontation with communist regimes ultimately became key actors in bringing down communism. The Polish anticommunist opposition was fueled by the Solidarność workers' movement. In Czechoslovakia and Hungary, civic groups consisting of academics, students, and common citizens allied with trade unions pushing to subvert communist rule. Paradoxically though, civic organizations were again largely marginalized after 1989-90, as the socioeconomic transformation process was generally spearheaded by technocratically operating executives (Dimitrov et al., 2006). Hence, scholars have argued that the It is frequently argued that interest groups face formidable pressures to professionalize their activities. The professionalization of lobbying skills enables organizations to more effectively cater to members' interests and gain access to decision-makers (Beyers, 2004). The training and deployment of lobbyists, the creation of internal monitoring systems, and accumulation of specialized knowledge can be seen as key investments in the viability and political clout of organizations. Hence, they have strong incentives to professionalize in order to enhance their advocacy capacities and position in policymaking (Albareda, 2020). In particular in new democracies, organizational professionalization can potentially be seen as an additional key to democratization, as the more professionalized organizations are, the more effectively they can interact with policymakers in the interest of constituents. Hence, professionalized organizations may ultimately fortify the linkages between decision-makers and various segments of civil society, that is the members of organizations and others benefiting from the collective goods they produce. Organizational professionalization may also facilitate the transmission of input and expertise 'from below' to policymakers.
Seen critically though, organizations with highly professionalized lobbying staff may also become detached from their member base (Hwang and Powell, 2009;Heylen et al., 2020), potentially resulting in declining membership. While professionalization is generally seen as being correlated with better access to policymakers (Albareda, 2020), it need not automatically enhance the democratic quality of organizations, as tedious consultations and consensus building with members may be cumbersome to efficient lobbying (Grömping and Halpin, 2019). 1 1 Albareda (2020) finds that groups directly involving members in decision-making are not more likely to access the EU Commission.
In either case, the question of professionalization is particularly interesting in the postcommunist context. Although it is shortsighted to speak of a complete tabula rasa for organized interests after 1989 , their playing field has radically changed. Postcommunist democracies faced unparalleled challenges in establishing effective institutions of political representation to channel civic demands into policymaking. A substantial growth in organizational foundations occurred, in particular during the democratization and EU accession process . Reform-oriented governments have also aimed to promote social dialogue, appear more responsive to citizens and manage social divisions by establishing new platforms for formal consultations with interest groups (Ost, 2000).
Altogether, democratization transformed the linkages between organized associations and the state, resulting in unprecedented new opportunities for organizations to develop. Unlike under communism, where interest intermediation was channeled through the state and party bureaucracy, CEE interest groups now operate within parliaments, ministries, and new interest intermediation forums, while also fostering ties with governments, oppositional parties, and like-minded partner organizations. They are free to calibrate their strategies to members' needs, consolidate ties with rivaling groups, or actively lobby the public outside formal political channels. Thus, there are strong reasons to assume that CEE interest groups are dedicating greater energy to the professionalization of their internal operations and strategic activities.

Professionalization of interest groups in CEE and beyond
Professionalization is a relatively marginalized area of interest group research, as most studies shy away from exploring the internal workings of individual groups due to enormous methodological challenges. In addition, the term professionalization is used somewhat elusively. For example, some authors explore how grassroots movements evolve into formal professional organizations (e.g., van Deth and Maloney, 2012). Other studies analyze the lobbying profession from a sociological perspective. Under the banner of 'learning to lobby', there is a large strand of research dealing with the personal qualities and identities of lobbyists (e.g., Coen and Richardson, 2009;Holyoke et al., 2015). Diverse researchers have looked at how lobbyists gather expertise and cultivate political ties (e.g., Groß and Kieser, 2006;McGrath, 2015). Another strand of research explores professionalization as an independent variable by testing how organizational professionalization impacts membership and political influence (e.g., Hollman, 2018;Heylen et al., 2020).
Yet how exactly do we conceptualize professionalization and what has the existing literature taught us? According to Imig and Tarrow (2001), professionalization is a process of transformation of internal structures of organized interests. This entails the creation of a professional workforce and culture, in which specialized knowledge is applied to respond to social problems. Professionalizing organizations thus aim to enhance the skills of their members through formal training, the development of monitoring and benchmarking systems, and more systematic reflection on organizational strategy. Hence, both 'operational' and 'expert knowledge' is necessary (Carmin, 2010: 187), which is usually 'acquired through the professional training of lobbyists' (Staggenborg, 1988;McGrath, 2005: 125).
So far, only a few studies comparatively explore the driving forces behind the internal development of interest groups, most of them covering only a few cases (e.g., Greenwood, 2002;Maloney, 2008;Saurugger, 2012;Fraussen, 2014). However, more recent studies by Klüver and Saurugger (2013), Maloney et al. (2018), and Czarnecki and Riedel (2021) have significantly advanced research. Klüver and Saurugger (2013) explore whether the professionalization of EU-level interest groups differs across group type. Accordingly, 'cause groups' often aim to transition from volunteer-based organizations towards organizations with a professional workforce. However, mainly focused on protest, they often lack fixed bureaucratic structures and routines. This makes them more susceptible to internal divisions and demobilization. They argue that sectional or business-oriented groups are more focused on members' special interests that generate concentrated costs and benefits, which in turn materially affect their supporters (Klüver and Saurugger, 2013). Generally, such groups can better monitor and mobilize their members and manage internal divisions. Moreover, as sectional groups primarily pursue insider lobbying (Binderkrantz et al., 2015), they may need to hire professionalized lobbyists and effectively monitor their success. Yet Klüver and Saurugger find no significant differences across group type, rather that financial endowment explains (at least a small share of) the variation in professionalization levels. Czarnecki and Riedel (2021) engage more systematically with Europeanization processes, conceptualized as both formal umbrella organization membership and more informal ties with likeminded organizations from other European countries as drivers for the professionalization of CEE interest groups. They find that it is not membership in supranational umbrella organizations or funding from EU-level interest organizations that necessarily lead to professionalization, rather, to a much greater extent, strong informal ties with organizations from other European countries pursuing similar objectives. Their findings somewhat contradict those of a panel-based study by Maloney et al. (2018) which identifies EU funds as a key driver of professionalization of approximately 50 Slovenian interest organizations.
However, we argue that these studies only paint a partial picture of the dynamics driving organizational professionalization anddespite their laudable advancementsperhaps insufficiently grasp the multiple international, domestic, and organization-specific catalysts for professionalization.

Theory, hypotheses, and variables
Our analysis advances preexisting research by integrating a more differentiated scope of potential determinants of professionalization and weighing their specific impacts against one another. We contend that three bundles of factorsorganizational funding sources, interorganizational cooperation, and (organizations' standing in) national interest group systemsmay be decisive. Our framework is based on our conviction that there are numerous potential pathways towards professionalization. Specifically, professionalization may be driven not only by 'hard' (i.e., monetary resources), but also 'soft' factors (i.e., interorganizational communication and learning) as well as the competitive environments in which organizations operate at the national level. In the following, we present our tri-dimensional framework, which builds on, but also adds new twists to preexisting analyses of the determinants of organizational professionalization (e.g., Maloney et al., 2018;Czarnecki and Riedel, 2021).

Organizational funding sources
First, it is intuitive that organizations require resources to professionalize (Baumgartner, 2009). Greater financial means allow groups to hire experts or invest in expertise accumulation, which can be used to lobby politicians. As highlighted above, Klüver and Saurugger (2013) determined that organizational resources are much more definitive for professionalization than organizational age, size, or type.
Yet not merely organizational resources per se, rather the source of funding may be decisive. Maloney et al. (2018) contend that another important factor for professionalization processes is public funding. Governments may effectively mandate groups to adopt specific organizational structures. As Czarnecki and Riedel (2021: 201) argue: 'public funding : : : may simply demand professionalization', as government subsidies may not only offer organizations a key resourcerelated advantage over non-subsidized groups, but also may require interest groups to adopt specific modes of operation or internal standards (see also Fraussen, 2014).
In the same vein, EU funds may be an additional catalyst for organizational professionalization. Generally, such funds are managed and distributed by member state governments themselves (Nicolaides, 2018), thus constituting a special form of government subsidies. In fact, Crepaz and Hanegraaff (2020) argue that many interest organizations would cease to exist without EU funding. Regarding CEE, Císař and Vráblíková (2010) showed that Czech interest groups receiving EU funds operate with significantly larger budgets, hire more staff, and develop their management structures, a finding echoed by Maloney et al. (2018) for Slovenian interest groups and Guasti (2016) for NGOs in CEE in general.
Hypothesis 1a: Organizations receiving subsidies from their own governments are more likely to professionalize.
Hypothesis 1b: Organizations receiving EU funds are more likely to professionalize.

Interorganizational cooperation
Besides potential new funding sources, EU accession has reshaped the playing field for CEE interest groups. The accession process brought about new national-level organized interests aiming to impact the negotiation process. Increasing ties with like-minded western organizations and EUlevel advocacy groups may have triggered processes of networking, learning, and diffusion, as EU and western organizations have provided substantial logistic and strategic support for various CEE civic organizations (Carmin, 2010;Guasti, 2016;Fagan and Kopecký, 2017;Vándor et al., 2017). Along these lines, European and international networking and cooperation between organizations may also function as a catalyst for organizational isomorphism, in particular mimetic isomorphism (DiMaggio and Powell, 1991). It is frequently argued that organizations seeking to secure their survival align themselves with structures perceived as successful and legitimate in their organizational environment. This is particularly relevant for CEE organizations, which historically suffer from numerous weaknesses in interest articulation, such as the frequent lack of clear strategies, skills, and organization, a lacking representative character and often underdeveloped channels for interest intermediation (Borragán, 2001: 64-67). The intensification of ties with western counterparts may prompt CEE organizations to mimic the structures and processes of foreign organizations perceived as 'more professional', which in turn may enable them to overcome their structural weaknesses.
Of particular significance in this regard are international umbrella organizations. Whether situated at the EU level or beyond, umbrella organizations bundle the interests of national-level members, enable national organizations to participate in transnational policymaking, and represent their interests internationally. As Eising (2016) shows, EU umbrella organizations provide member associations crucial information on EU policies and thus function as 'middle-men' between the national and supranational level. In line with hypothesis 1b, the proximity to EU institutions may also facilitate organizations' access to project funding. For example, Hanegraaff and van der Ploeg (2020) found that EU umbrella membership boosts the financial and professional resources of domestic groups, while also increasing access to EU-level institutions. Going beyond this, we argue that EU umbrella organizations may provide the additional long-term benefit of being an arena for the isomorphic emulation of organizational strategies perceived as successful (DiMaggio and Powell, 1991). Yet it is also equally plausible that transnational synergetic effects take place outside umbrella organizations and instead be stimulated by alternative forms of interactions such as bilateral expertise-sharing, coordination, and/or involvement in epistemic policy communities (Haas, 1992).

Hypothesis 2a:
Organizations which are members of European/international umbrella organizations are more likely to professionalize.
Hypothesis 2b: Organizations which network with like-minded European/international organizations are more likely to professionalize.
However, this transnational perspective potentially overlooks the fact that organizations ultimately operate foremost in national ecosystems and may equally benefit from interactions with like-minded domestic organizations. Recently, the interest group literature has shed light on interorganizational cooperation as a key determinant of lobbying success. Junk (2019), for example, shows that lobbying coalition diversity may signal broad social support for legislation, while a recent analysis of healthcare organizations by Horváthová and Dobbins (2021) reveals that cooperation between groups significantly enhances access to parliaments and governing parties. Yet national interorganizational cooperation may offer additional benefits to organizations regarding their internal development. While scholars have written extensively on the impact of transnational communication on policy emulation and convergence (e.g., Knill, 2005), we contend that national-level interorganizational cooperation may equally unleash isomorphic effects. Tight synergies with like-minded domestic groups, that is when coordinating strategies or sitting jointly on advisory boards, may make organizations increasingly aware of their own weaknesses, promote the share of expertise, experiences, and personnel, and subsequently unleash mutual mimetic and learning effects.
Hypothesis 3a: Organizations that are members of national umbrellas are more likely to professionalize.
Hypothesis 3b: Organizations that closely cooperate with like-minded national-level organizations are more likely to professionalize.
Self-critically, we could also argue that professionalization itself (i.e., time, staff, and resources) is a precondition for interorganizational cooperation, hence raising issues of causality. Nevertheless, we assumein line with mimetic isomorphismthat national organizational ecosystems generally comprise various more professionalized organizations which 'weaker' organizations aspire to cooperate with, learn from, and emulate, hence ultimately boosting their professionalization.
Organizations' standing in the domestic interest group system: competition and 'insiderness' Previous accounts of organizational professionalization also arguably overlook the domestic lobbying context. While the stringency of lobbying laws may affect professionalization (see Rozbicka et al., 2020), we argue, more broadly, that the competitive environments in which they operate may be an equally crucial determinant. In simple terms, organizations find themselves in a permanent struggle with other organizations for members, resources, and access to policymakers. As Hannan and Carroll (1992) maintain in their seminal analysis of organizational ecology, as organizational density grows, the 'supplies of potential organizers, members, patrons, and resources become exhausted'. In other words, the more organizations operating in a specific area, the greater the interorganizational competition. High organizational densities also limit interest groups' access to policymakers . We argue that the competitive environment may reverberate at the level of organizations (see Berkhout et al., 2021 for a recent study), as the quest for survival in dense environments and heavy competition from rivaling organizations may prompt them to optimize their structures and operations.
Hypothesis 4: Organizations that perceive stronger organizational density are more likely to professionalize.
Finally, we contend that insider status in the national interest group system may be a crucial predictor of professionalization. We follow Fraussen's (2014) argument that institutionalized interactions with public authorities may fuel organizational development, regardless whether these result from corporatist-style arrangements or less formally systematized interchanges. In an analysis of a Flemish environmental organization, he shows that its involvement in policy formulation and/or implementation enabled it to accrue material and immaterial resources to further invest in its internal development. Along these lines, professionalization may also be driven by the accumulation of expertise through intensive, state-interest group interactions, in particular, when organized interests participate in advisory boards. While it cannot be ruled out that exclusion from policymaking may also prompt groups to recalibrate their strategies (see e.g., Beyers and Kerremans, 2012), we argue in line with Fraussen (2014) that 'insiderness' is more likely to be a catalyst for professionalization.
Hypothesis 5: Organizations that perceive strong coordination with the state are more likely to professionalize.

Data and methods
Between February 2019 and June 2020, we conducted an online survey of national-level interest groups in healthcare, energy, and higher education policy in Czechia, Hungary, Poland, and Slovenia. The policy areas represent large portions of public budgets and are key to the long-term well-being of nations. Moreover, they are not interrelated, thus providing solid foundations for general insights. The country selection is motivated by their broader similarity (e.g., postcommunist democracies, new market economies, EU members), but also by diversity regarding three decisive characteristics for interest group politics: election funding, lobbying regulations, and the level of political-economic coordination/corporatism. Czechia has a very liberal economy with privately funded elections and weaker lobbying regulations (McGrath, 2008;Šimral et al., 2015). It is ranked low in classifications of corporatism (e.g., Jahn, 2016). Poland also is a relatively weakly coordinated market economy and generally ranked lower on the corporatism dimension (Jahn, 2016). However, elections are publicly funded, and extensive lobbying regulations and reporting requirements existdespite various loopholes (Kwiatkowski, 2016;Rozbicka et al., 2020). Hungary exhibits stronger market coordination (Duman and Kureková, 2012;Tarlea, 2017), whereby elections are publicly funded and lobbying activitiesat least formallytightly monitored, but since 2010 only loosely regulated (Laboutková et al., 2020;European Commission, 2020). Slovenia is the most coordinated and corporatist market economy in CEE (Bohle and Greskovits, 2012;Jahn, 2016). Regulatory controls over lobbying, party funding, and electoral campaigns are comparatively weak (despite heavier coordination by the 2010 anticorruption law, hence providing a polar opposite case to Poland, see Rozbicka et al., 2020). Before carrying out our standardized survey, we compiled datasets of 'population ecologies' of all interest groups operating in the three policy areas and four countries. The main sources were national online registries for civil society organizations. We used a harmonized set of keywords for each language and filtered out regional-level organizations and companies listed in the registries. 2 2 For Poland, the National Court Registry (Krajowy Rejestr Sądowy -KRS), for Hungary the court registry of civil society organizations (Civil szervezetek névjegyzéke), for Slovenia the AJPES registry (Agency of the Republic of Slovenia for Public Legal Records and Related Services), and for Czechia the registry of the Czech Statistical Office. In addition to the harmonized Our survey first requested information on membership structures (e.g., employees, volunteers, member firms, individual members, etc.). We then asked each organization about affiliations with European, international, and national umbrella organizations. Additional questions addressed the frequency of consultations between interest groups, parties, and governments, generally on a 1-5 scale (never, biannually, annually, monthly, weekly). We also asked about the perceived level of political coordination on 1-5 scales.
The third segment focused specifically on professionalization-related activities such as the training of lobbyists, fundraising, strategic planning, and human resource development. It also included questions about financial sources and financial stability. We received responses from 428 CEE interest organizations with an overall response rate of 33.9% (see online appendix).

Construction of variables
Instead of reinventing the wheel, we largely align our definition of professionalization with preexisting conceptualizations, in particular Imig and Tarrow (2001) and Klüver and Saurugger (2013). We conceive professionalization as a process through which organized interests hire professionalized staff, ideally with special training (see also Heylen et al., 2020) with the aim of creating a professional workforce (McGrath, 2005). This, in turn, enables them to focus on their strategic orientation and reflect more systematically on their short-and long-term operations, while evolving into responsive organizations. Besides a stronger focus on financial stability (e.g., through increased fundraising activities), another key dimension of professionalization is the accumulation of expert knowledge pertaining to both internal organizational operations as well as technical issues the organization deals with (see Carmin, 2010;Albareda, 2020). Finally, and importantly, professionalized organizations are generally equipped with mechanisms to monitor the effectiveness of their operations. 3 Table 1 provides an overview of all variables.
In line with this broader understanding of professionalization, we took two steps to construct our dependent variableorganizational professionalization. First, we calculated the organizationspecific mean score of five variables reflecting increasing professionalization processes: organizational and human resources development, fundraising, strategic planning, and evaluation of efficiency and effectiveness. The wording (i.e., increased) is rooted in our strong assumption that CEE organizations were largely unprofessionalized approximately 15 years ago due to lacking traditions of organized lobbying and the dominant position of national executives in policymaking (Goetz and Wollmann, 2001). 4 To what extent does your organization focus on the following activities as opposed to 10-15 years ago (or since its founding, if founded more recently)? (1) organizational development; (2) human resources development; (3) fundraising; (4) strategic planning; and (5) evaluation of efficiency and effectiveness (scaled 1-5, see Table 1). 5 Similarly to Klüver and Saurugger (2013), we included staff in our composite variable of professionalization, as we believe having lobbying staff significantly reflects organizational development. set of search terms, we systematically checked major interest mediation forums in every country and the membership in national umbrella organizations. We also checked whether an organization existing in one country (e.g., cardiology association) exists in the others. Although the national registries contain reliable data on active organizations, we nevertheless crosschecked every organization with Google searches to be sure they were truly active at the time of data collection (homepages, reports on activity in the press, press releases, etc.). 3 We unfortunately lack data on the involvement of members in internal organizational decision-making. 4 As we do not have panel data to measure changes in professionalization over time, we had to find an alternative measurement. Since professionalization is a rather subjective concept, we believe a question about current professionalization levels would not have delivered precise values. Thus, we opted for the wording "recent increase in professionalization" as a more accurate measurement. 5 A variable pertaining to training lobbyists was dropped due to the low response rate. We attribute this to the general aversion towards and negative connotation of the term "lobbying" in CEE (see Pospieszna & Vetulani-Cęgiel 2021). This aspect is supported by other studies which emphasize hiring professionalized staff, ideally with special training, as a key fundament of professionalization (e.g., Heylen et al., 2020).
Due to the extreme spread of staff numbers spanning from 0 to nearly 6000, we opted for a categorical variable (1-5) based on quintiles. Our aggregate variable for organizational professionalization 6 is thus constructed as the mean of the above-mentioned variables and a categorical variable for staff scaled 1-5. 7

Independent variables
We collected organization-specific data on the share (0/100) of government subsidies (H1a) and EU funds (H1b). EU funds are generally granted to national governments, which allocate them to specific organizations (Nicolaides, 2018). This makes a clear-cut distinction of funding sources difficult. Nevertheless, our survey questions asked for the specific share from both national Notes: Coding of professionalization: composite variable based on the row mean of focus on organizational development, human resources development, fundraising, evaluation of efficiency, and effectiveness as well as strategic planning (each measured on a scale 1-5 where 1 = much less than 10-15 years ago, 2 = less, 3 = same as before, 4 = more, 5 = much more than before) and the variable staff (scale 1-5 representing 5 quantiles based on the total amount of staff). Coding of state subsidies, EU subsidies: indicated as share of budget in percentages. Coding for membership in European and/or international umbrella: 0 = no member, 1 = member in either one European or one international umbrella, 2 = member in either multiple European and/or international umbrellas, 3 = member in either one/ multiple European and/or one/multiple international umbrellas, 4 = member in multiple European and international umbrellas. Coding of networking w. like-minded organizations: 1 = no, 2 = somewhat, 3 = very much. Coding of joint strategies: composite variable based on the row mean of cooperation in joint statements, joint political strategies, in representation on advisory boards (each measured on a scale 1 = never, 2 = occasionally, 3 = frequently). Coding of national umbrella: 0 = not part of national umbrella; 1 = part of national umbrella. Coding of the perception of growing number of interest groups: 0 = not growing, 1 = growing. Coding of perception of political coordination: 1 = very weak, 2 = weak, 3 = moderate, 4 = strong, 5 = very strong. Coding of concentrated/diffuse: 0 = diffuse, 1 = concentrated. Coding of age: age as of 2020 transformed into a logarithmic scale. Coding of perception of financial stability: 1 = stable for less than 1 year, 2 = stable for 1-2 years, 3 = stable for 3-5 years, 4 = stable for about 5 years, 5 = stable for more than 5 years. Source: own data and own elaboration.

6
Cronbach's alpha suggests a reliability of 0.7931 for the newly created index of professionalization, while the overall level Kaiser-Meyer-Olkin measure of sampling adequacy (KMO) is 0.7643. Factor analysis has shown that all components of professionalization except for staff have factor loadings between 0.5344 (focus on financial resources) and 0.7664 (evaluation of efficiency). For staff the factor loading is 0.3234 only. However, theoretical literature (Klüver & Saurugger 2013) suggests that staff is crucial for measuring professionalization. Hence, we include the variable in our definition. 7 A similar procedure was also applied by  who constructed their dependent variable based on the mean, hence creating a semi-continuous variable. governments and EU sources separately, and we assume that our respondents credibly distinguished between both original sources. 8 To grasp the effects of national and transnational interorganizational cooperation (H2a&b, H3a&b, respectively), we test umbrella membership and interorganizational cooperation separately both at the national and international levels. National umbrella membership is captured by a dummy variable (0 = no membership, 1 = membership). For the international level, we created a composite umbrella membership variable which entails whether an organization is part of a European or international umbrella organization. 9 Considering that transnational networking may occur outside umbrella organizations, 10 we also included variables capturing such cooperation. International networking is measured based on our survey question: 'In recent years, have you increasingly networked with like-minded organizations abroad when trying to influence national legislation?' on a 1-3 scale (no; yes, somewhat; yes, very much).
Assuming that active engagement with other national-level organizations may also stimulate professionalization, we constructed a composite organization-specific variable 11 consisting of three dimensions: 12 How often does your organization cooperate with other interest organizations in: (1) representation on advisory boards; (2) joint statements; and (3) joint political strategies? (never, occasionally, frequently).
We took two steps to generate variables reflecting groups' standing in the domestic interest group systems. Assuming that organizational density may enhance professionalization (H4), we asked: In your opinion, is the number of interest organizations attempting to influence decisionmaking and legislation in your area increasing, decreasing or stable over the past 10-15 years? (strongly decreasing to strongly increasing, 1-5). As for the perception of political coordination with the state (H5), we asked: How would you rate the level of policy coordination/political exchange between the state and your interest group? (very weak to very strong, 1-5).
We also controlled for the perceived levels of financial stability of interest groups (scale 1-5), group type, that is whether groups represent concentrated or diffuse interests, 13 and transformed organizational age into a logarithmic scale. 8 Mean VIF of the independent variables when conducting collinearity diagnostics is 1.18 and the condition number after centering 2.064. Hence, we assume no collinearity issues. 9 First, we asked our respondents whether they are members of a European and/or international umbrella organization -0=no membership, 1=umbrella membership, 2=multiple umbrella memberships. Second, we added up the values for membership in a European and international umbrella and created a variable with values between 0=no membership in either a European or international umbrella organization and 4=multiple memberships in European and international umbrella organizations. The coding for membership in European and/or international umbrellas hence is: 0=no member, 1=member in either one European or one international umbrella, 2=member in either multiple European and/or international umbrellas, 3=member in either one/multiple European and/or one/multiple international umbrellas, 4=member in multiple European and international umbrellas (see Table 1). 10 Another justification for this separate measurement of international networking is the fact that the term "umbrella organization" was unfamiliar to many of our respondents and that an established corresponding term is missing in some of the national languages.

11
For the index of joint strategies, Cronbach's alpha suggests a reliability of 0.7999 and the overall KMO is 0.7129. As the factor loadings are all above 0.7 and the Spearman's correlations all exceed 0.55, we stick to a measurement based on this index. 12 We dropped the variable grasping cooperation in fundraising due to the low number of responses, which reflects the lacking tradition of joint fundraising between CEE organizations. 13 We manually coded groups representing the interests of medical doctors and the medical and pharmaceutical industry, private hospitals and healthcare technology as concentrated. Concentrated energy groups included producers and associations of energy suppliers (fossil, nuclear, renewables), while concentrated higher education groups represent the academic profession as well as lobbyists for the university system. Diffuse organizations comprised patients', healthcare awareness and employees' organizations and student organizations. For energy, we coded consumers' groups, groups promoting clean and sustainable energy as well as energy efficiency and energy employees' organizations as diffuse interests.

Analysis
Are CEE interest groups professionalizing? Do they vary in terms of venues and strategies for professionalization?
Our descriptive statistics show that there is a striking lack of variation between countries, but indeed an increasing trend across the board (Figures 1 and 2).
Figures reflecting various moderate country differences regarding interorganizational cooperation, networking with European organizations, umbrella memberships, and external funding sources (i.e., EU and national subsidies) can be found in online appendix (Appendix Figures  A1-A5).
Our aggregated dependent variable for professionalization in the multivariate analysis is semicontinuous, hence enabling us to stick to a linear regression model with cross-sectional panel data (see also Klüver and Saurugger, 2013;. Country fixed-effects were applied to control for country-specific omitted variable bias. We first conducted separate regressions based on our three bundles of factors potentially driving professionalizationorganizational funding sources, interorganizational cooperation, and (organizations' standing in) national interest group systemsto single out the statistically most significant variables within these groups of factors. In a second step, we conducted two additional multivariate analyses (see Table 2), one including all proposed variables potentially affecting professionalization (Model 4), and one including only the variables that proved significant in models 1-3 (Model 5). We apply linear regressions as the fit was normally distributed. Values on the dependent variable are quite normally distributed with only one outlier, which is not problematic (Norman, 2010) (see Figure 2). The models explain up to 23.2% (M4) of the variance within and up to 78.8% (M3) between our panel units with an overall value up to 21.8% (M4 and M5). 14 In Model 1 (M1), we tested the effects of funding sources on professionalization, while also checking for the financial viability, age, and type of organizations (concentrated/diffuse). Unsurprisingly, our control variable financial stability has a relatively weak but significant positive effect on organizational professionalization. Put simply, the more financially stable an 14 The random effect models behaved similarly. organization, the more it can invest in further professionalization measures. However, the findings become more interesting when we differentiate funding sources. Although EU funds may be a major lifeline for CEE interest groups, our data also show no relationship between EU funding and increased professionalization (see Czarnecki and Riedel, 2021 for a similar finding). This lends support to an argument put forward by Guasti (2016) that many western/EU-based donors left CEE after approximately 2005-10 and shifted their activities further east (e.g., ex-USSR, Balkans) because the transformation process was seen as complete. State subsidies, however, have a rather weak, but significant and robust effect: a standard deviation increase in the share of state subsidies of the organizational budget enhances the professionalization composite index by 0.22 points (H1, see Table 1 for the summary statistics of the variables).
In Model 2, we tested to what extent interorganizational cooperation enhances professionalization. We did not find support for international (H2a) or national umbrella membership (H3a). This is surprising in view of other findings which show that umbrella membership can significantly enhance the political clout of interest groups (Fraussen et al., 2015). In contrast, networking with like-minded EU-level organizations and developing joint strategies with other domestic groups proved to be significant and of similar strength (M2). In the overall models (M4 and M5), this effect remains likewise strong. Similarly, the coefficient for joint strategies with domestic organizations has a positive effect on professionalization. Hence, we can confirm H3b: cooperation with like-minded organizations appears to promote organizational professionalization. Put differently, across our sample we find less formalized interorganizational cooperation activities such as networking with EU-level organizations and developing joint strategies with domestic groups to be more significant predictors of professionalization than formalized forums of cooperation such as umbrella organizations.
We now explore the set of factors grasped in Model 3, that is organizations' standing in the domestic interest group system. As shown above, we asked individual organizations about their perception of whether the number of interest groups was decreasing, stable, or increasing in their area. As shown in Table 2 (Model 3), the perceived increase in the number of organizations proved significant for organizational professionalization and shows a robust effect throughout the tested models. This is an important finding because it shows that interest groups adjust their strategies to  the 'environmental pressures', that is, to increased competition for resources, constituents, and influence at high organizational density levels. There is also a strong relationship between perception of being incorporated into the political coordination process and professionalization, that is every unit increase on the five-point scale reflecting political coordination is associated with 0.169 unit increase in professionalization. However, this latter effect loses its strength and significance in M4 and M5. Finally, our control variable 'concentrated/diffuse' in Models 1 and 3 shows that being a concentrated interest correlates with not increasingly professionalizing. Instead, it appears to be more diffuse groupshere patients' and healthcare awareness groups, ecological associations as well as employees' and students' organizationswhich are investing in internal organizational development. However, the effect of 'concentrated/diffuse' proves to be heavily dependent on model specifications once we control for interorganizational cooperation and the domestic interest group system (M2, M3, M4, and M5). We also estimated all models with random effects instead of country-fixed effects, which yielded similar results. Hence, we believe our findings to be robust (see online appendix Table A1).
However, the demonstrated relationship between political coordination with the state and professionalization evokes the classic 'chicken-and-egg problem': does professionalization itself enhance political coordination with the state, or is the state actively promoting the professionalization of certain organizations? We believe the former is more plausible. As noted above, we also find state subsidies to be a predictor of professionalization, albeit relatively weak. Nevertheless, this lends some evidence to the side of the equation that it is often ultimately the state driving the professionalization of a certain set of organizations through funding incentives. This somewhat supports Olejnik's (2020) argument on 'patronage corporatism' in the region, namely that postcommunist governments often find themselves in a symbiotic relationship with certain politically favorable organizations. Specifically, governments engage in political consultations and exchange with social partners with coinciding interests, many of which often directly receive government subsidies to accumulate political support. Our results reveal that such privileged organizations may capitalize on their proximity to the state by professionalizing their internal structures.
In Model 4, we test the robustness of our findings including all variables. What is interesting is that, among the funding-related variables, only the share of state subsidies and financial stability stay significant. The variables testing international networking, as well as the perception of political coordination all prove to be robust, although it is again difficult here to determine the direction of causality, that is whether being professionalized is a 'ticket' to transnational networking activities, or whether transnational networking is the driver of professionalization. The same applies to M5, in which we included only the variables that proved significant in Models 1-3. However, some effects are slightly weaker here. Yet the effect of national-level joint strategies with other organizations remains significant and relatively strong in M5 as does the perception of political coordination and networking.
Thus, our data essentially reflect three storylines. First, as shown with different degrees of significance in Models 3-5, we see that the domestic interest group system makes a difference. An increasing number of groups in the playing field is prompting organized interests to professionalize, hence suggesting a link between density and professionalization. Second, Models 4 and 5 add weight to the observation about patronage corporatism (Olejnik, 2020): although the effect is relatively weak despite statistical significance, organizations benefiting from state subsidies and actively coordinating with the state appear to be funneling resulting material and nonmaterial resources into further professionalization, which in turn further consolidates their symbiotic relationship with the state. The state in turn may draw on such organizations to garner public support for its policy agenda, likely to the detriment of excluded organizations. On the upside though, the beneficial effects of joint action for achieving policy goals with domestic peers and networking on the EU/international level also are supported by our data. That is, interest groups lacking patronage/clientelistic embeddedness can potentially compensate their disadvantage with intensive domestic and international cooperation. In other words, networking and joint strategymaking can unleash mimetic isomorphic effects between organizations and ultimately boost the professionalization of previously structurally weaker organizations.

Conclusions
While this analysis allowed us to present a crucial first step in mapping the professionalization patterns of CEE interest groups, the lack of time-series data makes it difficult to gauge the validity of the causal mechanisms over time. 15 However, also in view of fluctuating political landscapes, our approach has enabled at least preliminary insights into the multitude of factors driving the internal professionalization of interest groups in CEE. First of all, our findings underscore those of Klüver and Saurugger (2013) and Maloney et al. (2018) that organizational resources are key to professionalization. However, unlike Maloney et al. (2018), we found EU funding to be less important for professionalization than support from national governments. In line with Fraussen (2014) and Olejnik (2020), our data indeed show a 'visible hand' of the state in promoting organizational professionalization.
Moreover, our results lend much support to the soft power of isomorphic learning, both through domestic interorganizational joint strategy-making and transnational networking (outside formalized umbrella organization structures). Finally, our analysis is the first to our knowledge to tentatively link organizational professionalization with organizational density (conceptualized as the perception of growing numbers of groups). This paper has the disadvantage of having aggregated data from four specific countries for the sake of a larger number of observations. Future research should therefore zoom in on the causal drivers of organizational professionalization within specific countries over time. Moreover, we encourage researchers to develop more exact measurements of population density and more systematically assess its effect on professionalization. Scholars should also further pursue the already developing research agenda on the relationships between professionalization, (hierarchical vs. bottom-up) internal organizational decision-making, and membership (Albareda, 2020;Heylen et al., 2020;Berkhout et al., 2021). The question of internal organizational democracy (or the lack of) is particularly relevant in the CEE context where interest organizations were historically appendages of the party-state bureaucracy and lacked traditions of directly incorporating members into decision-making. Finally, and importantly, the presented data reveal little about the effects of selective state funding on singular groups. Specifically, are (national-conservative) CEE governments driving the professionalization of a diverse array of rivaling organizations (e.g., patients' and medical associations, employer's and employee's associations, 'clean' and 'dirty' energy organizations) or only subsidizing certain segments of the organizational landscape, hence unilaterally driving professionalization? (e.g., only 'dirty' energy or employers)? Future analyses should therefore definitely explore how the ideological couleur of both governments and organizations affects the flow of state funds and which types of interest groups are benefitting.