This article was downloaded by: [193.84.56.46] On: 12 February 2014, At: 06:37 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T3JH, UK Democratization Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/1oi/ fdem20 Promoting competition or cooperation?The impact of EU funding on Czech advocacy organizations Ondřej Císař3 & Jiří Navrát i Ib a Department of Political Science, Charles University, Prague, Czech Republic b Department of Public Economics, Masaryk University, Brno, Czech Republic Published online: 10 Feb 2014. To cite thisarticle: Ondřej Císař & Jiří Navrátil , Democratization (2014): Promoting competition or cooperation?The impact of EU funding on Czech advocacy organizations, Democratization, DOI: 10.1080/13510347.2013.869742 To link to thisarticle: http://dx.doi.org/ 10.1080/13510347.2013.869742 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the "Content") contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions Democratization, 2014 http://dx.doi.org/10.1080/13510347.2013.869742 Promoting competition or cooperation? The impact of EU funding on Czech advocacy organizations Ondfej CisaP* and Jifi Navratilb "Department of Political Science, Charles University, Prague, Czech Republic; bDepartment of Public Economics, Masaryk University, Brno, Czech Republic {Received 23 May 2013; final version received 30 October 2013) What impact does European Union (EU) funding of advocacy organizations have? To address this question our article turns to the post-communist Czech Republic, an ideal laboratory for studying externally dependent nongovernmental organizations. Employing social network analysis, the main objective of this article is to analyse the effect of EU funding on the cooperation networks of Czech advocacy organizations. Our source of data is a survey of these organizations. We aim to figure out whether there is an association between the dependency of advocacy organizations on EU resources, and their cooperation with other organizations. Contrary to the prevailing interpretation based on the competition argument, our hypothesis is that the greater the dependency on EU funding, the greater the cooperation capacity on the part of advocacy organizations. Keywords: NGOs; European Union; civil society; Czech Republic; democracy promotion Introduction According to the received wisdom, while foreign-driven civil-society-building programmes have contributed to the organizational capacity of individual advocacy non-governmental organizations (NGOs) in East-Central Europe, these programmes are believed to be much less successful in promoting their mutual cooperation. Assessing advocacy organizations in Poland, Ekiert and Kubik identify poorly coordinated Western assistance as a facilitator of "a considerable fragmentation within Polish civil society".1 As a result, they observe that "across all sectors of the newly reconstituted civil space there emerged (...) serious fragmentation, political divisions and intense struggle for resources ... "2 Similarly, focusing on Polish animal rights organizations, Jacobsson concludes that the existing funding context that makes them "compete not only for money but also the attention of the public (...)> is not conducive of inter-organisational * Corresponding author. Email: ondrej.cisar@ff.cuni.cz ©2014 Taylor & Francis R Routledge Taylor &. Francis Group 2 O. Císař and J. Navrátil cooperation".3 Due to foreign funding, cooperation networks of advocacy NGOs in East-Central Europe are perceived as rather underdeveloped, especially in terms of collective action and coordination capacity.4 Drawing on data from the Czech Republic, and contrary to the prevalent interpretation which views external, most importantly EU, funding as a factor hindering cooperation, we claim that it promotes inter-organizational networking and cooperation among advocacy NGOs. In contrast to the old EU member states, Central-East European countries (the Czech Republic included) present a different context for what effect EU money can have on advocacy NGOs.5 These countries democratized relatively quickly in terms of formal rules and institutions, but lagged behind the established democracies in terms of civic and political activism: This weakness of civil society so far in post-Communist Europe is the most serious deficiency in the democratisation process there, since it suggests a lack of real depth which comes from (...) associational networks. For this reason, outside assistance to civil society has been all the more crucial...6 From the point of view of EU and other Western donors, a vibrant civil society has been understood as an important precondition of functioning democracy in the post-communist countries. In the era beginning shortly after the regime's collapse, citizens showed little interest in politics and little support for advocacy NGOs. Thus, the resources needed for their organizational survival had to come from foreign funding, and this is still a widespread phenomenon in the region.7 In the first half of the 1990s, local organizations were mostly supported by state and non-state agencies from the US and individual West European countries. Since the second half of the 1990s, European funds have become the most important source of external funding. Therefore, we focus on their impact in this article. Here the Czech Republic represents a broader group of Central-East European post-communist countries that have experienced a massive influx of EU funding in support of local political and social activism in the form of advocacy organizations. In this respect, the lessons learned in this context can be consequential for EU strategies in other democratizing regions, too. So far, however, only a limited amount of systematic empirical evidence has been gathered on Central-East European NGOs and the way their networks interact with the EU. In order to meet its goals, the article first attempts to define and operationalize the NGOs' cooperation capacity, and then to figure out whether it displays different values for the various thematic industries within the advocacy component of Czech civil society. The article focuses on advocacy NGOs; that is, groups that "make public interest claims either promoting or resisting change that, if implemented, would conflict with the social, cultural, political or economic interests or values of other constituencies or groups",9 and operationalizes their cooperation capacity as the central position within their cooperation network. Further and most Democratization 3 importantly, this paper asks whether there is a relation between network centrality and Czech advocacy NGOs' dependency on EU resources. NGOs and their networks There has been an enormous increase in the number of formally registered advocacy civil society organizations (understood as NGOs10) in post-communist Europe, including the Czech Republic, since 1989.11 In addition to the registered organizations, there are many unregistered informal groups and networks of activists, groups operating outside the legal requirements such as anarchist and Trots-kyite groups, or even secretly operating right-wing extremist, radical environmental, or anti-fascist groups. Nobody disputes that there has been an increase in the number of advocacy NGOs in post-communist settings: formally registered NGOs are seen as the organizational building blocks of civil society. They provide citizens with a number of goods, not only in the realm of advocacy, but also service provision. However, they are generally believed to be less capable of serving as a platform for 12"" broader (networked) social movements. Since their level of mutual coordination is low, their capacity for collective action is limited.13 Consequently, what we observe in post-communist countries are isolated organizations unable to cooperate among themselves and provide citizens with an organizational platform for collective action. According to available studies, instead of mutual cooperation towards the goal of mobilizing citizens and aggregating their demands and contributions, local NGOs became dependent on foreign sources of funding.14 A recent review of these studies summarizes the effect of foreign donor support as "competition, rivalry and uncooperative behavior among civil society organizations, which led to the fragmentation of civil society".15 Taking stock of Czech civil society in this journal, Fagan concludes that "NGOs compete with each other for scarce donor funding, they switch from project to project in response to the whims of donors ... "16 However, an alternative, yet still minor, approach claims that some of these organizations seem to fare much better than others in terms of their capacity to cooperate with other organizations. Tarrow and Petrova17 have even come up with a new label - transactional activism - to conceptually capture the cooperation of these groups. The notion of transactional relations is generally used to capture the ability of an actor to engage with other relevant (collective) actors through the exchange of control over physical or symbolic media.18 Specifically, transactional activism means "the ties - enduring and temporary - among organized non-state actors and between them and political parties, power holders, and other institutions".19 Although transactional activism is associated with cooperation among advocacy organizations, it is in fact a broader concept than advocacy, since it includes not only explicit attempts at issue advocacy, but all types of transactions including exchange of information and resources, coalition building around different projects, and collaboration with NGOs across thematic industries. 4 O. Císař and J. Navrátil In this respect, Tarrow and Petrova focus on what Diani and Baldassari have labelled transactions or "weak organizational ties" largely based on resource exchange and synonymous with inter-organizational cooperation, and not on "social bonds" based on shared membership.21 In fact, focusing on the post-com- 22 munist countries, Tarrow and Petrova seem to draw unwittingly on the well-established resource-mobilization-based argument about the substitution for "thick mobilization infrastructures" with their "thin" versions by social movements unable to tap into existing pools of potential followers.23 Although this article does not share McCarthy's focus on the mobilization of individuals, it nevertheless points in the same direction when it analyses the (thin) organizational infrastructure of specialized advocacy organizations, which seem to form the backbone of a significant part of contemporary political and social activism in post-communist settings. Since we prefer the concept of (inter-organizational) cooperation to its transactional synonym, cooperation is used in the remainder of the article. The effect of EU patronage on domestic organizations According to the widely shared consensus in the literature, political and social activism in post-communist Europe was deeply influenced by international (including EU) civil-society-building programmes, designed to help democratic civil societies develop in the target countries. These programmes, "which have promoted civil society development, have chosen to equate this with furthering NGOs".25 As a result, they created a system of international patronage that supplied advocacy NGOs with the necessary resources to maintain themselves despite rather unsup-portive local populations.26 While in the first half of the 1990s these resources were typically provided by US and US-based private foundations, with individual European states and foundations also playing an important role, by the end of the decade these sources began to dry up.27 In the second half of the 1990s the EU took over as the primary source: The prospect of EU accession (...) meant that the civic sector across the region felt the pull of Brussels more keenly than Washington ... As a result, civil society organizations in East Central Europe adopted a much more intense engagement with EU priorities in order to meet the conditions for accession ... 28 While there is a clear consensus on the influence of international patronage, its effects are rather disputed. This debate mirrors the generally established debate on the effect of external funding on advocacy groups' and social movements' autonomy and ability to act within the context of domestic politics. Whereas one group of authors views external patronage as a disempowering force for advocacy groups, taking away their autonomy and making them dependent on their donors, the other group either does not observe any effect, or sees an empowering influence instead.29 Unlike these contributions, which focus more generally on the advocacy Democratization 5 capacity of NGOs, this article specifically concentrates on the problem of organizational cooperation. According to one interpretation, international funding contributes to competition among NGOs, leading to fragmentation and isolation among the NGOs, and making them dependent on international donors and separating them from domestic activist networks.30 According to this argument, international patronage can help establish and maintain individual NGOs, but is unable to contribute to building their cooperation networks: "Rather than facilitating horizontal networks among groups, foreign aid strengthened the division of the civic community between the haves and the have-nots ... "31 The result is inter-organizational competition. An opposing argument claims that foreign-funding dependency actually results in inter-organizational networking and an increased ability to act on the 32 part of the supported organizations. The result is inter-organizational cooperation. We will elaborate on these expectations below. The competition thesis 33 According to theories on civil society development, by the beginning of the 1990s the civil society concept had become fashionable among democracy-promoting agencies around the world. In an idealized neo-Tocquevillian understanding of civil society promoters, civil society began to be seen as an associational arena formed by voluntary associations independent of both the state and the economy. Civic associations were expected to provide citizens with a vehicle for political participation, and to put a check on the decision-making processes taking place within the formal structure of state institutions. It was believed that by supporting these relatively independent NGOs, political reforms in transition countries such as the Czech Republic would be served better than if money were provided directly to agencies within the state bureaucratic structure.34 In other words, civil society promotion programmes were expected to help democratic tran- 3 5 sitions in post-communist countries. There was one more important reason for the EU to take active part in diffusing the civil society ideology: civil society has been expected to help the EU solve its "democratic deficit" problem.36 Support for civil society has been conflated with support for advocacy NGOs.37 According to the competition argument, by providing financial support the EU not only directly influenced the agendas pursued by these NGOs, but also redirected their activities from domestic cooperation and mobilization of their constituencies to grant-seeking.38 As a representative of a Czech human rights advocacy group stated: I would say that we focus more on the authorities, not on people ... because if you want to work somehow, you have to get the money... you can only get the money from Europe, or from the government, or from the regional government, or from the city or local government... so for us it is important to get the money and with the money I can realize my agenda ... I can do almost nothing without the money ... and it is the authorities that decide on the distribution of the money, not people ... 39 6 O. Císař and J. Navrátil According to Fagan, "the whole process of acquiring EU-derived funding has become more complicated, intensive and competitive for NGOs".40 Likewise, focusing on women's groups in Poland, Hungary, and Russia, McMahon claims that: "Foreign interest in helping women has thus unintentionally contributed to the (...) intense competition among groups with similar goals."41 Since funding is limited, NGOs must compete for access to authorities and institutions, and 42 receive grants at the expense of the others. Hence they prefer vertical to horizontal ties. Scarce resources are the reason they tend to guard their grant application know-how even after the money is granted, since they will need it again when applying for resources next time. This hinders their willingness to cooperate with other organizations and share valuable information. Thus instead of cooperation, external funding creates dependent organizations that engage in communication and resource exchange with their donors, but not with their domestic peers, whom they see as their competitors. This competition mechanism leads to fragmentation. HI: The more NGOs depend on EU funding, the less they cooperate. The cooperation thesis This article likewise argues that by signalling to local groups that formally registered and professionally managed organizations are most likely to receive funding, the EU has indeed helped to solidify an activism based on formal NGOs.43 At the same time, however, the article claims that civil society building programmes helped create organizations willing and able to engage in cooperative ties. Bruszt and Vedres44 make a similar argument that there is a positive relation between exposure to EU funding and the ability of the recipient organizations they studied (not only NGOs) to act autonomously; Císař and Vráblíková45 focus on the EU's ability to foster transnational protest by Czech advocacy NGOs; while Stark, Vedres and Bruszt46 challenge the fragmentation interpretation of transnational influence by showing a positive relation not only between all types of transnational ties and the ability of NGOs to act, but also the ability to mobilize individuals and establish coalitions across various sectors of society. Focusing specifically on Czech women's groups, Císař and Vráblíková47 observe a positive effect of EU influence on cooperation among local NGOs. To our knowledge, these are the only studies focusing on post-communist countries to explicitly argue that there is a positive relation between external resources and either the ability of NGOs to act or, as in case of Stark, Vedres and Bruszt and Císař and Vráblíková, to 48 cooperate. On a more general level, Jenkins49 argues that funding agencies' goals are complex; therefore, it is unsubstantiated to see them as agents breaking down coordination and cooperation among non-state actors. Similarly, Chavesc, Stephens and Galaskiewicz50 have demonstrated that external funding does not suppress coordination and advocacy activity among NGOs; on the contrary, external funding either has no effect or actually activates NGOs. In the case of the EU, a Democratization 7 positive effect of EU funding was demonstrated on pan-European cooperation of West European advocacy NGOs,51 especially among those advocating minority 52 and women's rights. In the Czech context, international funding has been the only viable strategy for advocacy NGOs to ensure their existence, especially in the fields of human and women's rights, since it was very difficult for them to establish resonance with the local population. For example, in the field of minority rights the overall national climate can be regarded as non-conducive in terms of potential support of both the state and the general public.53 It was EU funding that helped these groups to gain some autonomy and ability to establish cooperative ties. Drawing on the contributions of Stark, Vedres and Bruszt, and others, which are relatively unorthodox in the general context of research on the EU's impact in post-communist Europe, this article expects EU funds to have a positive effect on inter-organizational cooperation among advocacy NGOs. By supporting domestic organizations, the EU directly contributes to their capacity to act, which also entails mutual cooperation. In order to be able to meet the requirements attached to EU money, domestic groups must fulfil a wide range of formal criteria, and accommodate these requirements especially regarding dissemination and addressing multiple audiences. According to the projects' distribution requirements, the projects' achievements must be disseminated among various publics through conferences, workshops, and seminars.54 Networking is a vital tool for NGOs in achieving this goal. EU money, together with its guidelines, lead to increased pressure towards cooperation among recipient organizations. Therefore we expect that a rising proportion of EU resources in the NGOs' budgets will be accompanied by a positive effect on their key network attributes. H2: The more NGOs depend on EU funding, the more they cooperate. Controls Since this article is interested in the impact of EU funding dependency, it controls for other factors derived from theory as possible determinants of inter-organizational cooperation. These factors are individual membership, number of issue areas an organization is involved in, years of existence, and size of budget.55 Since it is clear that the ability of NGOs to cooperate varies across thematic industries, we include industry controls in our analysis.56 The first is the number of individual members an organization has. Drawing on Tarrow and Petrova,57 who see cooperation-based activism as a substitute for activism based on individual participation, we might expect organizations with smaller memberships to network more than their mass-based counterparts, which can rely on mobilized individuals. According to this expectation, organizations unable to mobilize individuals tend to substitute for this weakness by establishing cooperative ties with other organizations. 8 O. Císař and J. Navrátil Second, the number of issue areas an organization is active in is expected to increase the likelihood of (inter-industry) cooperation, since the more areas the organization focuses on, the more potential contacts with various partners it has. Third, we control for the number of years the organization has been in existence, assuming that the longer an organization has been in existence, the more cooperative ties it will have. Older organizations have simply had more time to establish cooperative relations. Fourth, we control for the size of an organization's budget, as this may play an important role in the centrality of an organization within the inter-organizational network. Drawing on resource mobilization theory, one can argue that the more resources an organization has, the more attractive it will be for other organizations seeking a solvent partner. Fifth, we control for the possible influence of the membership of an organization within the respective issue industries. Previous research indicates that post-materialistically oriented organizations in areas such as the environment and human rights display more cooperative predispositions than materialistically 58 oriented groups such as trade unions ; therefore we must control for industry membership in order to isolate the possible effect of EU funding dependency we are primarily interested in. Data and measurement Since the main dependent variable in our study is cooperative relations among different NGOs, we have chosen social network analysis (SNA) for data collection and retrieval. Acknowledging the critical importance of careful sampling of nodes and operationalization of ties within network analysis,59 we distinguish several basic aspects in our approach. Generally there are three basic steps to identifying a network: identification of nodes, identification of ties, and identification of network boundaries. We apply this scheme through two successive but partially independent steps: first, we identify the set of nodes; second, we identify the ties, and finally the network boundary. First, with an NGO as our basic unit of analysis (node), we opted for a snowball sampling. The snowball question ("Please name groups or organizations that belong to the same movement or industry") was not limited to any particular number of organizations. In order to reconstruct and map as much complexity in the relationships as possible, the snowball started simultaneously with five key organizations (selected according to expert opinion) in all of the thematic industries of organized Czech civil society shown by previous research to be the most active. These were human rights and gay/lesbian and women's rights, environmental, trade union and agrarian interests, social assistance, and underground radical Left activism. Only groups and organizations mentioned at least twice are included in the sample; the response rate was 70%. The survey was carried out over the period of October 2007-December 2008. For the purpose of this analysis we further narrowed our node sampling according to our research question. Democratization 9 Accordingly, we included only advocacy NGOs, that is, organizations in the environmental, women's, and human rights industries, trade unions, and agrarian organizations. In all, 101 core organizations were selected (25 environmental organizations, 29 women's rights groups, 24 human rights organizations, 14 trade unions, and nine agrarian organizations). In order to supplement our knowledge of how EU funding dependency promotes cooperation among NGOs, and to understand in more detail the possible mechanism^), two waves of interviews were carried out. The first wave of interviews was done in the same period as our organizational survey (more precisely between March 2006 and September 2008): 13 advocacy organizations in the Czech Republic and nine NGO platforms in Brussels were interviewed in the fields of environment, human and women's rights, and good governance. Second, we supplemented this knowledge with two additional interviews in December 2012 with the project manager of a human rights organization and an interview with a fundraising specialist working for several Czech environmental organizations. In order to assess various levels of potential EU funding impact, we opted to interview persons in different relevant organizational positions with long-term experience in the field. Respondents were asked about their experience with the context of application and implementation of various types of EU funding schemes, and with their consequences for organizations and their external relations with other groups, institutions, and the public. The interviews lasted between one and three hours. Dependent variable The focus of our study is on cooperation capacity; we define this generally as the capacity of an organization to enter into and maintain horizontal exchanges with other subjects. Consequently, two organizations were said to have ties when an organization was listed as cooperating by the interviewed group. To fully capture the multifaceted notion of cooperation among NGOs, representatives of interviewed organizations were asked the following question: 'According to their importance, name groups, organizations or networks that your group cooperates, consults, or communicates with." Furthermore, respondents were asked to evaluate each of the listed relationships on a one-five scale according to their intensity and frequency. All listed cooperative ties were substantiated by the respondents (name, type, and example of cooperation) in order to avoid vague and non-specified inter-organizational relations. The final cooperation network was symmetrized (method average; that is, the tie between two organizations has the average value of two unidirectional ties between these organizations) and contains 636 nodes and 1032 ties. We operationalize the cooperation capacity of an NGO as its centrality within the overall cooperation network. In other words, the more an NGO is able to cooperate with other organizations, the stronger and more numerous cooperative ties it has and the more central it becomes within the whole structure of cooperation. To operationalize the concept more precisely, however, one needs to consider several possible 10 O. Císař and J. Navrátil measures of network centrality at the node level. Specifically, these measures are all-degree centrality (that is, the number of both incoming and outgoing ties of a node, in-degree representing the incoming ties and out-degree representing the outgoing ones); or the betweenness centrality measure, indicating the ability of a node to be the bridge on the shortest lines between many other nodes and thus exert control over the information flow. There are two basic problems with these measures. The former reflects only immediate ties and activities with adjacent nodes, and ignores the potential indirect impact (influence) beyond them, while the latter reflects a too-specific dimension of the flows within the network, related more to the control of information flow but downplaying other potential and indirect exchanges within the network. Another measure of node centrality is closeness centrality, which reflects the distance between the organization and all the other NGOs within the network.61 These measures are conceptualized quite differently than the betweenness measure; they primarily represent the capacity of a node to escape the control potential of other nodes (because of the existence of many alternative ties), and to quickly interact with others.62 This conceptualization is much closer to the notion of NGO cooperation capacity, defined as the capability to become embedded within a multitude of cooperative ties with other NGOs; and also as the potential to quickly enter into cooperation with new actors. For the purposes of our study, however, we specifically opted for the eigenvector centrality measure, which adds another desired characteristic to the basic closeness centrality approach: it can also be used for valued networks, bringing in the dimension of quality (intensity) of cooperative ties. As with the basic closeness centrality perspective, eigenvector centrality mirrors the overall structure of the network, but this time through weighting neighbouring contacts according to their own centralities.63 In other words, it is a function of the centralities of its adjacent vertices, and the vertices that are adjacent to these vertices, and so on. The higher the number and quality (value) of ties that the organizations cooperating with a particular organization have, the higher the value of eigenvector centrality for this organization.64 This measure reflects the capacity of an NGO to be effectively engaged in multiple-intensive exchanges in the cooperation network, to maintain a privileged position with regard both to adjacent and more distant organizations, and to be tied to other important and privileged (qualitatively and quantitatively well-connected) organizations. The eigenvector centrality of each organization was computed in UCINET Software.65 Independent variables EUfunding dependency: We use the proportion of EU funding in an organization's budget that was given by the interviewed representative of an organization, and based on its official records. Having no EU funding in the budget is coded as zero. Missing values (non-responses) were completed from the annual reports of the respective organizations. Democratization 11 Number of issues: Representatives of organizations were asked which of the listed issue areas they are involved in (civic rights and freedoms, agriculture, education, housing, transportation, social services, energy, poverty, Third World development, environment, fiscal policies, information policies and media, consumer protection, health care, art, regional development, sport, minorities, equal opportunities, globalization, corruption, activity of government or political parties, defense and foreign policy, and other). A positive answer was coded as one, and all positive answers were counted. Years of existence: This variable is measured as the difference between the answer to the question "In what year was your organization founded?" and the actual year. Individual membership: We measure this variable as the answer to the question: "How many members does your organization have?", after representatives declared that the organization is based on individual membership. Budget: Budget is indicated by the question on the total income/size of the NGO's budget in 2006. Because there are some missing data (specifically, trade union organizations refused to answer this question), the values of this variable were recoded so as not to lose the number of cases. The new coding consists of 10 values that represent 10 groups of equal size. A qualified estimate on the basis of publicly available annual reports was made in order to measure the cases with a missing value on this new 10-value scale. The variable was controlled against the number of full-time employees, and it yields the same results. Descriptive statistics for all variables are shown in Table 1. Controls for industry: Each representative was asked to select the first, second, and third most important label for their organization out of 25 options (labour, agrarian, environmental, women rights, human rights, minority rights, disabled people's rights, GLBT (gay, lesbian, bisexual and transgender), anarchist, communist, Marxist, Trotskyite, global justice, nationalist, religious, socially oriented, feminist, peace, cultural, leisure time, sport, professional, development, education, political, other). We categorized membership in an industry on the basis of the first choice. An organization was included in an industry if it described itself as primarily focused on the area and if it was mentioned at least twice by other organizations that declared themselves as primarily focused on the same area. Analysis To conduct ordinary least squares (OLS) regression analysis we used the node-level regression command in UCINET.66 It is based on an algorithm that enables an analysis of inter-dependent observations by recalculating the regression coefficients over a large number of iterations (we set the level to 35,000) with random redistribution of covariates among the unit of analysis while maintaining their network structure. ^-values are computed as a proportion of permutations with statistics as large (or as small) as those computed at the beginning.67 Table 1. Descriptive statistics: five industries. Labour Human rights Women's rights Environmental unions Agrarians NGOs NGOs NGOs TV 14 9 24 29 25 EU funding (%) Mean 0.69 5.00 41.18 34.43 29.16 Std. deviation 1.49 9.14 27.64 36.68 22.48 Individual membership Mean 27,581 348 27 949 329 Std. deviation 34,988 657 81 4272 1104 Number of issues Mean 6.57 4.11 7.29 5.24 5.40 Std. deviation 3.16 1.69 2.44 2.06 2.40 Years of existence Mean 18.64 16.00 12.54 13 18.84 Std. deviation 3.88 5.10 4.58 7.67 6.21 Budget in 2006 (total in EUR) Mean 1,524,444 318,250 289,796 118,507 224,342 Std. deviation 1,496,990 236,502 196,282 135,477 234,191 Missing values 5 1 1 1 0 Eigenvector centrality Mean 0.25 0.29 9.66 7.85 13.66 Std. deviation 0.67 0.29 9.41 7.40 10.86 Source: Survey of Czech NGOs. Democratization 13 Cooperation networks in the Czech Republic We begin our analysis with a brief description of NGOs in five selected thematic industries. To compare the attributes of the organizations that will be included in our explanatory model, we focus on individual membership, number of issues, the age of the organization, budget, and their eigenvector centrality (Table 1). Although as expected the average labour union membership far exceeds the others, it is rather unexpectedly followed by women's rights groups (which are usually thought to be small in the Czech Republic), and with agrarian organizations closer to other post-materialist (the environment, human rights) groups. This is due to two things: first, there is one outlier in the women's rights industry with a substantial membership inherited from the previous regime; second, agrarian groups are primarily concerned with lobbying (and service provision) instead of mobilization. All the industries are similar in their number of issues, which is usually around six. On average, the oldest organizations are - not surprisingly - in the labour and environmental industries, as they continue in their traditions established well before 1989. The distribution of eigenvector centrality reveals that there is a clear and significant gap between the cooperation capacities of post-materialistically and materialistically oriented organizations (Table 1). While environmental, human, and women's rights NGOs tend to occupy a privileged central position within the system of cooperation with plenty of ties to other centrally positioned NGOs, the agrarians and labour union organizations are rather isolated and find themselves on the periphery of the network. It seems obvious that post-materialistically oriented groups tend to occupy a more central position within the cooperation networks of Czech domestic groups. Not only have they achieved a better (more central) position than the agrarians and trade unions, but they also enjoy a considerably higher capacity to establish ties across their issue industries and to maintain interactions with other well-positioned organizations, thus further improving their cooperation capacity. On the other hand, the organizations from two materialistically oriented industries not only have scarcely any contacts outside the industry, but their internal pattern of cooperation is not very dense nor intensive, and sometimes (with the trade unions) even takes the form of a star network. In the next step we explore the relation between use of EU resources and the issue area of domestic organizations. The resource variable is operationalized as the percentage of EU resources in the budget of a particular organization. The share of EU finances for individual industries is shown in the first row of Table 1. The basic pattern is clearly visible: while there is no doubt that environmental, women's rights, and human rights NGOs are largely dependent on EU funding, the situation is just the opposite in the case of trade unions and agrarian organizations. Finally, we focus on the main theoretical argument of this article and analyse the relationship between EU funding dependency and cooperation capacity: what are the differences among sampled issue industries? Are there visible patterns 14 O. Císař and J. Navrátil Figure 1. EU funding dependency and cooperation (spring embedding layout). Source: Survey of Czech NGOs. Note: The size of a node stands for the value of its eigenvector centrality. The nodes' shapes stand for their industry membership: a square denotes labour unions; rounded square: agrarians; circle: environmentalists; upwards triangle: human rights; and downwards triangle: women's rights NGOs. The colour denotes the share of EU resources in the budget: white stands for 0-20%; light grey: 21-40%; medium grey: 41-60%; dark grey: 61-80%; and black: 81-100%. Only NGOs under study are displayed; the strength of the ties is not depicted. of dependence between the share of EU money in an organization's resources and the capability of that organization to establish and maintain horizontal ties with other NGOs? Figure 1 essentially gives us a positive answer. First, it seems apparent that NGOs with a large proportion of their funds from EU sources tend to be in a good central position within the cooperation network, and are characterized by intensive ties to other well-positioned NGOs (that is, with a high level of eigenvector centrality). Even if some of the most central NGOs do not rely entirely on this type of funding, they regularly display a considerably higher share of it. On the other hand, only a handful of organizations that use little or no EU resources display a higher cooperation capacity. The EU effect on cooperation capacity The figure presented in the previous section suggests that there is a relationship between resource access and the properties of organizational networks. Organizations that mobilize resources internally tend to be located on the periphery of the cooperation network; EU resource dependency seems to correlate with a more central positioning of organizations within the network. In order to test Democratization 15 this hypothesis statistically we conducted an OLS multivariate analysis, controlling for the variables commonly associated with network centrality: number of individual members, number of issue areas an organization is active in, number of years the organization has been in existence,68 budget, and membership in the respective thematic industries. As specified in the theoretical section, we include EU funding dependency as the independent variable hypothesized to have either a positive or negative effect on the cooperation capacity of NGOs. The model in Table 2 includes all the hypothesized variables as explanatory factors of cooperation capacity. Controlling for "traditional variables", the model suggests that external EU funding dependency helps groups engage in cooperation: the more organizations rely on EU resources, the more they are connected in inter-organizational networks. This result supports our second - cooperation - hypothesis at the expense of its competition counterpart. Compared to the "traditional variables", EU funding dependency is by far the most powerful determinant. Also, our additional data analysis suggests that the proportion of EU funding in organizations' budgets does not produce competition among them. Logistic regression analysis (results not presented in a table) shows that EU funding is not related to NGOs perceiving themselves in a situation of economic competition with other NGOs, controlled for the same variables as in our core analysis. In our analysis we controlled for size of membership, number of issues, length of existence, size of budget, and the membership of an NGO in a particular industry. First, in line with what was hypothesized, there is a significant negative effect of membership on cooperation: the fewer the members, the more organizations cooperate. A possible explanation might be that organizations without a large membership compensate for their inability to mobilize individuals through inter- Table 2. EU funding dependency and cooperation (OLS regression results, TV = 101). B St. B EU funding dependency 0.071*** 0.223 Controls Individual membership -0.000*** -0.017 Number of issues -0.023*** -0.006 Years of existence 0.057*** 0.040 Budget 0.372*** 0.094 Controls for industry Labour unions 12.143*** 0.453 Agrarian industry 11.939*** 0.368 Women's rights industry 5.463*** 0.267 Human rights industry 4.812*** 0.221 (Constant) -60.306*** 0.000 Adj. R2 0.230 Source: Survey of Czech NGOs. Notes: /^-values determined by permutation tests. *p < .05; **p < .01; ***p < .001. Number of iterations: 35,000. 16 O. Císař and J. Navrátil organizational cooperation with other NGOs. Second, although we expected a positive relationship, issue multiplicity has instead shown a negative effect on the eigenvector centrality of the sampled actors. A possible explanation might be the clear preference by most organizations for cooperation within particular issue industries that was shown by our network analysis (not displayed here); this diminishes the importance of dealing with a wide array of issues in order to be connected. Third, the age (length of existence) of an organization has the hypothesized positive effect. Fourth, the budget of NGOs has a presupposed positive effect on cooperation. Uniformly, the effects of control variables on cooperation are much weaker compared to our main variable, EU funding dependency. The significant effect of EU funding dependency exists even after controlling for industry membership. In regard to controls for industry membership, our network analysis seems to show considerable differences among particular industries in terms of the positions of their members within the network, intensity of cooperation, and the structure of their resources. Indeed, the model revealed that the network centrality of NGOs was considerably affected by their industry membership. Discussion and conclusions The presented findings are in line with our cooperation argument, according to which EU funding dependency has contributed to cooperation and networking among Czech advocacy NGOs. Regarding our research on EU funding dependency, the competition argument put forward by many researchers of post-commu-nist civil societies does not seem to hold. Czech organizations that are dependent on external resources tend to cooperate more than organizations not dependent on external funding. All in all, our analysis shows that EU resource dependency increases the likelihood that advocacy organizations will network among themselves and engage in mutual cooperation. This challenges the still-prevailing understanding of the effects of external funding on NGOs and collective action in general. Our article shows that EU funding dependency can hardly be seen as a factor hindering inter-organizational cooperation and the potential development of broader social movements in the Czech Republic.71 On the contrary, it contributes to the establishment of such network structures among local NGOs. In addition, thanks to our network analysis we are able to identify the types of organizations that engage in cooperation while being dependent on EU funding. The post-materialistically oriented NGOs show much higher values in this respect than their materialistically oriented counterparts. Since the competition argument was formulated mostly in the studies of organizations in the environ-mental and women's rights industries, our finding is particularly striking. While the more EU-dependent environmental, women's, and human rights NGOs tend to network more, the opposite holds true for the less EU-dependent trade unions and agrarian organizations, which rely instead on other institutional Democratization 17 sources, on their members, and their own activities. Consequently, the visualization of the relationship between share of EU resources and eigenvector centrality is fairly persuasive (see Figure 1). At the same time, we lack survey data from different time periods to validate our findings longitudinally and establish a dynamic relationship between EU funding and a consequent change in the level of NGOs' cooperation. However, we have qualitative data convincingly tracing mechanisms that lead from the reception of EU money to NGOs' cooperation. According to our interviews, the need for NGOs to obtain grants forces them to focus on strategic planning and cooperate with other NGOs in order to achieve the prescribed goals of the project, which are often aimed at multiple policy areas and audiences. Thus informal inter-organizational ties are typically brokered during work on EU-funded projects73; in other words, they are established after the application of an NGO is approved. Few direct official project partnerships among NGOs are strategically organized until funding has been granted. Domestic project partnerships are not usually formally supported during the process of application assessment, and applicants have even been explicitly warned against choosing unreliable project partners by the authorities themselves.74 This is actually the exact opposite of the competition mechanism leading to fragmentation. According to the present mechanism, while scarcity of resources may lead to non-cooperation in the preparatory phase, that is, before the grant is awarded; the situation fundamentally changes after funding is granted and the project is launched. In this phase, the need to gain additional expertise from other organizations and to establish partnerships with them to successfully meet the project's requirements and disseminate its results outweighs competitive pressures.75 In a nutshell, upon the disbursement of the money, competition gives way to a search for partners. The second identified mechanism leading to increasing cooperation among NGOs operates in a rather indirect manner. The conditions of EU-funded projects lead to substantial capacity building, professionalization, and institutionalization on the part of NGOs, which subsequently improves both the quality of services they provide and their capacity to promote these services; thus their prestige and position within civil society "significantly improves".76 This contributes to being perceived by other NGOs as useful partners in potential coalitions and cooperation networks, and increases their network attractiveness to potential partners. To wrap up: in this article we have focused on networks among Czech advocacy organizations in order to see whether there are differences regarding cooperation among various advocacy NGOs. The article has indeed identified important differences among more networked post-materialistically oriented NGOs and less connected materialistically oriented organizations. The main finding of our article shows that contrary to the established belief, external funding from EU resources contributes to increased cooperation among NGOs, resulting in the establishment of instrumental organizational ties. 18 O. Císař and J. Navrátil Effectively, the article has shown that EU funding dependency, measured by the proportion of EU funding in domestic groups' budgets, does not lead to competition/fragmentation of the national networks of advocacy organizations, but to their integration through intensified cooperation. While it has been shown elsewhere77 that the EU in a neo-pluralist fashion supports underrepresented interests, especially the environmental, human, and women's rights advocacy groups and so-called social NGOs, at the European level, our analysis seems to demonstrate that it helps consolidate the cooperation networks of this type of advocacy organizations on the Czech national level as well. Funding This work has been prepared as part of the research project Collective Action and Protest in East-Central Europe (code GAP404/11/0462) funded by the Czech Science Foundation. It was also supported by the project "Employment of Newly Graduated Doctors of Science for Scientific Excellence" (CZ. 1.07/2.3.00/30.0009) co-financed from European Social Fund and the state budget of the Czech Republic. Notes 1. Ekiert and Kubik, "Civil Society From Abroad," 55. 2. Ibid., 35. 3. Jacobsson, "Fragmentation," 367. 4. Baker and Jehlička, Dilemmas of Transition; Carothers, Aiding Democracy Abroad; Fagan, Environment and Democracy; Fagan, "Taking Stock"; Guenther, "The Possibilities and Pitfalls"; Howard, The Weakness of Civil Society; Jacobsson, "Fragmentation"; McMahon, "Building Civil Societies"; Mendelson and Glenn, The Power and Limits; Narozhna, "Foreign Aid." 5. Tarrow and Petrova, "Transactional and Participatory Activism"; Bernhagen and Marsh, "Voting and Protesting"; Pridham, Designing Democracy. 6. Pridham, Designing Democracy, 204; see also Howard, The Weakness of Civil Society. 7. Fagan, Environment and Democracy. 8. But see Bruszt and Vedrés, "Associating, Mobilizing, Politicizing"; Stark, Vedres, and Bruszt, "Rooted Transnational Publics." 9. Andrews and Edwards, "Advocacy Organizations," 481. 10. See Carothers, Aiding Democracy Abroad. 11. See Fric et al., "The Czech Republic." 12. For example, Fagan, Environment and Democracy; Fagan, "Taking Stock"; Guenther, "The Possibilities and Pitfalls"; Henderson, Building Democracy; McMahon, "Building Civil Societies." 13. Jacobsson, "Fragmentation," 354; Navrátil, "Between the Spillover." 14. Carothers, Aiding Democracy Abroad; Fagan, "Taking Stock"; Guenther, "The Possibilities and Pitfalls"; Henderson, Building Democracy; Howard, The Weakness of Civil Society; McMahon, "Building Civil Societies"; Mendelson and Glenn, The Power and Limits; Narozhna, "Foreign Aid." 15. Jacobsson and Saxonberg, Beyond NGO-ization, 7. 16. Fagan, "Taking Stock," 534. 17. Tarrow and Petrova, "Transactional and Participatory Activism." 18. Knoke and Yang, Social Network Analysis, 12. 19. Tarrow and Petrova, "Transactional and Participatory Activism," 79. Democratization 19 20. Baldassarri and Diani, "The Integrative Power." 21. Diani, "Social Movements," 227. 22. Tarrow and Petrova, "Transactional and Participatory Activism." 23. McCarthy, "Pro-Life and Pro-Choice." 24. Ibid. 25. Pridham, Designing Democracy, 204; see also Carothers, Aiding Democracy Abroad. 26. Aksartova, "Why NGOs"; Baker and Jehlička, Dilemmas of Transition; Císař, "Externally Sponsored Contention"; Fagan, Environment and Democracy; Fagan, "Taking Stock"; Fagan and Carmin, Green Activism; Guenther, "The Possibilities and Pitfalls"; Howard, The Weakness of Civil Society; McMahon, "Building Civil Societies"; Nar-ozhna, "Foreign Aid." 27. Fagan, Environment and Democracy. 28. Stark, Vedres, and Bruszt, "Rooted Transnational Publics", 330; also Dimitrova and Pridham, "International Actors"; Lavenex and Schimmelfennig, "EU Democracy Promotion." ■^t- 29. See Brooks, "Is there a Dark Side"; Brown and Troutt, "Funding Relations"; Chavesc, o Stephens, and Galaskiewicz, "Does Government Funding"; Froelich, "Diversification ^ of Revenue Strategies"; Gazley and Brudney, "The Purpose (and Perils)"; Jenkins, g "Channeling Social Protest." 2 30. Carothers, Aiding Democracy Abroad; Ekiert and Kubik, "Civil Society From Abroad"; Fagan, Environment and Democracy; Fagan, "Taking Stock"; Howard, ^ The Weakness of Civil Society; Jacobsson, "Fragmentation"; McMahon, "Building ^ Civil Societies"; Mendelson and Glenn, The Power and Limits; Narozhna, "Foreign co Aid"; this is sometimes captured by the concept of NGO-ization; see Guenther, "The Possibilities and Pitfalls"; for a critical review see Jacobsson and Saxonberg, Beyond NGO-ization. ^ 31. Henderson, "Selling Civil Society." 32. Bruszt and Vedres, "Associating, Mobilizing, Politicizing"; Císař and Vráblíková, in "Transnational Activism"; Stark, Vedres, and Bruszt, "Rooted Transnational Publics." ^ 33. Carothers, Aiding Democracy Abroad; McMahon, "Building Civil Societies"; Nar-