Networks of Interlocking Leaders: Exploring the Links and Identifying the Elites in four European countries

Abstract

This chapter aims at analysing the inter-organisational networks emerging from interlocking leaders among the most resourceful, national-level civil society organisations (CSOs) in four European countries. By comparing four different country contexts, we analyse differences in the characteristics of networks, find different mechanisms that seem to structure organisational links across the cases, and identify who the most powerful actors are based on the CSOs’ network positions. While we find small components of tightly connected organisations in the Italian and Polish cases, mainly around similar policy areas, these contexts are characterised by overall fragmentation. In Sweden and in the UK, on the other hand, we find that there are a handful of key CSOs that link the majority of the resourceful organisations, thus occupying powerful network positions.

Introduction

A myriad of Civil Society Organisations (CSOs) of various sizes and characteristics constantly interact with each other, for instance in building alliances and coalitions for policy work, collaborating for public campaigns, and coordinating for mobilisation of the masses (protests, petitions, etc.). The organisational ties established by top-level leaders with multiple organisational affiliations are crucial in that they mediate interpersonal connections and interactions among the leaders of different organisations who are involved in strategic decisions (Willems et al., 2015). Yet, the leaders interlocking multiple CSOs through cross-representation in decision-making bodies, such as boards, have seldom been studied when it comes to the question of power and influence in civil society (see a recent review by Yoon, 2020).

While there is a view of CSOs as ‘diverse, highly specialised and horisontally integrated organisations’ (Messamore, 2021), recent scholarly work has highlighted increasing hierarchisation in terms of resource concentration, as well as formalisation and centralisation within the institutionalised field of civil society (Johansson & Uhlin, 2020; Santilli & Scaramuzzino, 2021; Scaramuzzino & Lee, forthcoming). By studying the networks of interlocking top-level leaders in the most resourceful CSOs operating at the national level, this chapter addresses the question of who the power elites are in the organisational field of civil society in four European countries and how we can understand the connections between them.

We aim at identifying the most powerful actors in the networks of CSOs at national level, and also across different national contexts. By comparing four different countries we also want to understand differences in networks across contexts. An underlying assumption in our approach is that the individual leaders who are in leading positions in more than one civil society organisation are conceived as central agents that facilitate information exchange and transmission of knowledge between the organisations (Haunschild & Beckman, 1998). The information and knowledge, as well as the personal ties that are established through these agents by holding multiple positions in different organisations are deemed to generate strategic advantages for the organisations’ capacities to be influential within the field of civil society as well as vis-à-vis external actors (Granovetter, 1985). This perspective is in line with elite theory where access to resources, occupying strategic positions, and operating in networks of influence are all considered as sources of social power (Domhoff, 2002; López, 2013; Mills, 1956; Yamokoski & Dubrow, 2008).

Previous studies looking specifically into the phenomenon of leader interlock in the civil society sector have found empirical evidences for the positive impact of board interlocks among CSOs and the likelihood for organisational collaborations (Guo & Acar, 2005; Ihm & Shumate, 2018), and that for instance such collaborative relations can lead to better organisational performance, such as accessing larger public grants (Faulk et al., 2017; Paarlberg et al., 2020). Others have found that organisations relying on similar funding sources, of similar sizes, and sharing similar operational activities, are more likely to be connected via interlocking boards (Willems et al., 2015).

In our study, we approach the networks of interlocking leaders among CSOs with a more agnostic position as to the reasons why the interlocks exist. We argue that there could be other than strategic reasons (i.e., that the interlocking leaders can lead to better organisational performance via networks) for the leader interlocks to emerge among CSOs. It could for example be historical or ideological bonds among certain organisations that lead to leader interlocks (Messamore, 2021), or simply by chance, through interpersonal networks where people invite others to be part of a board based on personal trust and confidence or people having interest and being engaged in multiple issues.

Once they are formed, however, the existence of multiple interlocking leadership positions among major CSOs could function as a way of coordinating a given civil society field, a field populated otherwise by a wide range of organisations with diverse characteristics. Studying the organisational links established through leader interlocks at the top level can therefore tell us something about how a civil society field is structured and who the most powerful actors are, by means of occupying central positions in a network. Moreover, studying interlocking leaders can help us identify which organisations and individual leaders have strategically favorable positions, for instance by occupying broker positions.

This chapter aims at addressing the following research questions by studying a cross-sectional picture of organisational links among CSOs from a network perspective: (i) What network structures do we observe among the interlocking top-level leaders of the most resourceful CSOs in Italy, Poland, Sweden, and the UK? (ii) Who are the most powerful actors based on their network position?

The top-level leaders in our study include the board members, chairs (including vice-chairs), and also the executive leaders (e.g., CEOs, chief directors and their deputies). As to the first question all four country contexts are included, while regarding the second research question we delve into the two country contexts where we observe giant component network structure (i.e., a connected component of a network that includes a significant proportion of the entire nodes in the network) of interlocking leaders among CSOs: Sweden and the UK. The four national contexts included in the study allow us to explore similarities and differences across the contexts.

The chapter is structured as follows. After the introduction, the next section includes our argument and purpose for country selection, the methodology used in identification of the most resourceful CSOs, the description of data for interlocking leaders, and the analysis methods used. The chapter proceeds with a result section consisting of a first part summarising basic network structures of the Italian and Polish cases, where we do not observe any giant components, and a second part where we identify the most central actors occupying the power positions in the networks in the Swedish and the UK cases. In the concluding section, we summarise the findings and discuss possible interpretations of the findings across countries.

Data and Method

Country Cases and Sampling Elite Organisations

Our study explores interlocking leaders between resourceful CSOs at national level in four different countries: Italy, Poland, Sweden, and the UK. These four countries have been associated with different ‘civil society regimes’ in previous research (e.g., Salamon & Sokolowski, 2018). Italy, an example of Continental or corporatist regime, is mostly service-oriented with a large share of paid staff. Poland, as an example of the Eastern or post-communist regime, is also service-oriented, yet with smaller workforce and with a very small share of paid staff. Sweden, as an example of a Nordic or Social democratic regime has a mostly advocacy-oriented civil society sector with a relatively large workforce mostly made up of volunteers rather than paid staff. The UK, as an example of Anglo-Saxon or liberal regime, is characterised by the prominent role of civil society as service provider, with a large proportion of paid staff (Archambault, 2009; Salamon et al., 2017; Salamon & Sokolowski, 2018).

This regime-typology has been developed comparing the role of the civil society sector in each country and predominant resources that the organisations are equipped with. The extent to which the interlocking leaders among CSOs and the networks they produce would differ following different civil society regimes is not self-evident. The organisations in our sample, as will become evident in the following, are not representative of the whole civil society sector in each country. They are resourceful organisations in the sense that they control disproportionately large amounts of resources, both economic and political ones. They are also national-level organisations involved in coordination of regional and local actors, political representation of segments of civil society vis-à-vis the national government, as well as in service and capacity-building activities involving their members and constituencies.

For a systematic comparison, the populations of CSOs in all four countries have been identified and delimited by a set of indicators of financial and political resources, while considering the contextual specificities. The organisations were identified through a series of systematic screening procedures for each country, using the indicators measuring different types of financial and political resources according to the Multi-dimensional Measure of Resource Stratification in Civil society (MMRSC) (Scaramuzzino & Lee, forthcoming, see also Appendix to the volume).

Based on this sampling procedure, different internal structures in terms of coordination and resource stratification within the communities of national elite organisations in each country tend to appear. The Italian and Swedish elite organisations follow a pattern of different levels of coordination with many umbrella organisations representing organisations active within specific policy areas and even representing the whole civil society sector. This “Russian doll” structure corresponds to higher resource stratification with a few organisations controlling many types of resources and the majority controlling fewer resources. A less hierarchically structured pattern of coordination is observed in Poland and the UK, with fewer networks and umbrella organisations. This less hierarchical structure corresponds to a pattern of resource stratification with no organisations (or very few as in Poland) controlling all types of resources used in our method (Scaramuzzino & Lee, forthcoming; see also Appendix).

Mapping Boards and Interlocks

Based on the systematic mapping method of CSOs (see Appendix, this volume), we identified 293 national level CSOs for Italy, 447 for Poland, 394 for Sweden, and 434 for the UK. In the next step, we collected the names of leaders occupying the top-level positions in the identified organisations, such as board members, chairs, and also the executive leaders. The data was collected in 2019 for Italy, Sweden, and the UK and in 2020 for Poland. The following table (Table 10.1) presents the sampling of data and the number of CSOs and leaders that were found to be interlocking different organisations in each context. Some patterns are evident. In Sweden and the UK, a substantial share of the organisations are interlocked by the leaders; 48% of the organisations for the UK and 40% for Sweden. For Italy, only about 28% of organisations were connected via interlocking leaders while for Poland even less, 17%. It is also relevant to notice that the share of leaders interlocking different organisations among all identified leaders is even smaller, ranging from about 5% of the leaders for Sweden to 2% for Italy.

Table 10.1 Multiple affiliations in four countries

Multiple affiliations among the top-level leaders of the most resourceful CSOs that we identified in each country context is in other words a phenomenon concentrated to a rather small clique of the civil society leaders, yet in the case of Sweden and the UK involving nearly half of all identified CSOs. In the UK we find a particular type of leader that is not present in the other contexts. They tend to have the role of ‘ambassadors’ or ‘patrons’, not least including some individuals in the British Royal Family occupying positions in the boards of ten or more organisations.

Multiple affiliations that a given individual leader has in more than one organisation are the basis of the data for bipartite networks for each country, entailing two types of nodes (organisations and leaders). From this we create one-mode networks, where the links between the organisations via interlocking leadership positions become the main data.

Analytical Strategy

Identification of power elites or leadership groups, whether local or national, has traditionally followed one of four distinct strategies: the positional, the decisional, the reputational, and the relational approach (cf. Hoffman-Lange, 2017; see also chapter by Santilli and Scaramuzzino in this volume). In this chapter we employ the K-core algorithm, a method used to identify groups that can be considered elite in a relational perspective using social network analysis.

The relational approach draws on the notion of social circles to find a central circle of actors within the elite. The circle is identified by prominent members naming others as key partners, thus allowing the inclusion of power brokers. Inclusion in the central circle of these power networks is viewed as an indicator of the power structure and membership of the elite social circle. For example, this method was used in the identification of Danish elites (Grau Larsen & Ellersgaard, 2017), where they constructed a list of state organisations, parliamentary circles, NGOs, corporations, and foundations which were connected through participation in events and used a special weighted version of the k-cores algorithm, taking into account the relationship values between individuals and comparing the integrative effect of their different heterogeneous affiliations.

In this chapter, alongside other network centrality measures we employ the K-core algorithm in order to identify the organisations occupying the most powerful positions in the networks of interlocked leaders in Sweden and the UK. K-core measure is frequently used in identifying power elites in studies of elites (Corominas-Murtra et al., 2014; Huijzer & Heemskerk, 2021; Young et al., 2021). We apply this approach for studying the UK and Swedish cases in our study, where we can identify substantial network structures emerging from interlocking leaders among the CSOs. The K-core algorithm locates parts of the graph that form sub-groups such that each member of a sub-group is connected to a given number of the other members. That is, groups are the largest structures in which all members are connected to all but some number (K) of other members. Each individual is assigned a ‘coreness score’ corresponding to the minimum degree of individuals they are connected to. By decomposing an entire component, or progressively removing individuals with the lowest degree until further removal of individuals from the component leads to a decrease in the minimum degree, we eventually arrive at the core group. For example, to construct the four-core of a network, one first eliminates all nodes with three or fewer connections; this in turn leaves some nodes with fewer than four connections, so the process is iterated until those that remain have at least four connections each.

Besides the K-core algorithm we employ centralisation measures such as ‘degree centralisation’. We employ this measure to demonstrate the degree of internal cohesion and top-down integration of actors. Centralisation ‘measures the dispersion of centralisation scores relative to the most central score in the network’ (Sinclair, 2011, p. 30). According to this notion, a star-shaped network is the network with the most unequal degree of centralisation for any number of actors. In such a network, all actors except the central actor have a relationship degree of one, and the central actor has a relationship degree equal to the number of all actors minus one. In the operationalisation adopted here, we use this understanding of centralisation to demonstrate the degree of organisational connections through leader interlocks. We use also the measure of ‘betweenness centrality’ (Freeman et al., 1991). Centralisation of this type is based on the assumption that the actors will use all the links that connect them, proportionally to the shortest paths between organisations. The coefficient of centralisation of a given actor within the network is measured by the proportion of each pair of actors in the entire network (i.e., flowing through the shortest paths) (Borgatti, 2005). In addition, ‘eigenvector centralization’ (a measure of the influence a node has on a network) will tell us about the extent to which a network is dominated by a single node (Borgatti et al., 2018, p. 184). We use normalised measures to be able to compare the two country contexts.

For Italy and Poland, instead of employing the K-core approach and the network measures introduced above we opt for focusing on qualitative commentaries in order to understand the relatively fewer ties we find among the organisations via leader interlocks. We explore possible mechanisms behind the observed organisational connections through leader interlocks, focusing specifically on policy areas in which the organisations are active.

Results and Analysis

The comparative analytical lens through which we study the networks of interlocking leaders of CSOs provides an opportunity for understanding the different extent to which a given field of national CSOs is consolidated (better connected) or fragmented (loosely connected). As it turns out, among the four countries we include in our study, we observe relatively loosely connected networks of CSOs linked through interlocking leaders in Italy and Poland, and relatively densely connected networks in Sweden and the UK. Only 4% of organisations are connected in the Polish case (Connectedness = 1 minus proportion of pairs of vertices that are unreachable), while 8% of organisations in the Italian case are connected. When it comes to Sweden 47% of all organisations in our sample are connected and 63% in the case of the UK. We therefore apply partly different approaches in understanding the networks of interconnected CSOs among our empirical contexts. We start first with the Italian and Polish cases.

Italy

Looking at the network of Italian CSOs (Fig. 10.1), we find some components of organisations. In the following, we comment on the four components that have at least eight organisations linked to each other and try to explain these links focusing on policy areas, membership-based relations, and cultural/ideological affinity.

Fig. 10.1

Network of interlocking leaders in Italy. (Source: Authors’ own analysis)

One of the components is clearly connected by the common policy area of “international cooperation” which is how international aid and development often is framed in Italy. Central for this component are two organisations: AOI and Focsiv. They are both umbrella organisations but AOI is at a higher organisational level. Hence, Focsiv is member of AOI. Focsiv is the umbrella organisation for CSOs working with international cooperation and that have a common cultural/ideological point of reference in the Catholic movement. In fact, all CSOs linked to Focsiv (except for AOI) are members of this umbrella organisation. On the other side of AOI we find another umbrella organisation for international cooperation, namely COCIS, which is the umbrella organisation for CSOs that belong to the secular post-communist tradition. Also Forum SAD is a member of AOI while CeSPI is not. However, CeSPI as a study centre for international politics is involved in international cooperation and linked to a humanitarian organisation like Amnesty. Another organisation in this component is MCL which shares with its link CEFA the common Cristian catholic culture/ideology.

Another component includes FISH, an umbrella organisation for disability organisations, and the disability movement linked with the civil service organisations and the volunteering organisations. FISH is a central actor here with nine links which can be understood in different ways. Some organisations are clearly representing people with specific disabilities: AISLA, UILDM, FAIP, AIPD, and ANFASS. Many of these organisations are also members of FISH. Three other links can be understood in terms of common policy area or interest. The Theleton Foundation is funding research on rare diseases which would be relevant for FISH. ID is an organisation for control and accountability of the non-profit sector and in particular concerning private donations. Forum Nazionale Servizio Civile coordinates the civil service that conscientious objectors do instead of the military service. Many of these objectors are traditionally involved in volunteering. In this constellation, we also find CSV net, an organisation involved in organising volunteers. Forum Nazionale Servizio Civile is the other central actor in this component with five links (including FISH and one of its members). Linked to this organisation we find a sports organisation like OPES, and two other organisations mobilising volunteers and civil service, MODAVI and AMESCI. In the same constellation we also find an umbrella organisation for youth organisations, Forum Nazionale dei Giovani.

Another cluster revolves around the cooperative movement with Legacoop as the central organisation. Legacoop is an umbrella organisation for the cooperatives traditionally linked to the secular post-communist tradition. Many of the links are with organisations for the cooperative movement’s different sectors: a member of Legacoop such as ANCC-COOP (for consumer cooperatives), an organisation for the Promotion of the Culture of Co-operation, and of Nonprofit (AICON) research center for the cooperative movement (IRIAD), and lastly an organisation that promotes cooperatives using properties confiscated from the organised crime (Libera). In the same constellation we also find the organisation for political and trade unionist representation of the cooperative sector (Federsolidarietà Confcooperative). Linked to Libera we also find Gruppo Abele, an organisation working against drug addiction and social exclusion with strong historical ties to Libera. In the same constellation we also find an institution for studies of peace and disarmament.

The remaining larger component of Italian CSOs seems to be held together by a common focus on culture, tourism, and hobby. A central organisation holding together two smaller components is UNPLI an umbrella organisation for local associations for development, culture, and tourism. Linked to UNPLI we find an organisation for culture, tourism, and sports (FICTUS), which in its turn links to an organisation for culture and sports (AICS) and another for tourism (CTS). To UNPLI we also find a link to the civil service movement with organisations such as CESC, CNESC, and ARCI Servizio Civile as well as the volunteering with AVIS for volunteers for blood donors.

Poland

The network of Polish organisations via leader interlocks is the most fragmented, compared to the other three contexts in our study (Fig. 10.2). There is also a tendency to homophily at the global level of the entire network in terms of policy areas, and it is the only statistically significant result among our cases. In relation to policy area categories there are more of them, but Environment, Sport, Coops are very much responsible for the effect, as some of them are simply in ‘diads’, which means that if one node from a given category is connected with another node in the same category and there are no other nodes from the same category, we have 100% homophily. It concerns mainly Disability and Health, Environment, Sport, Coops, Religious organisations, and to a lesser extent Ethnic and Cultural organisations.

Fig. 10.2

Network of interlocking leaders in Poland. (Source: Authors’ own analysis)

The Polish network thus resembles many archipelagos of islands of organisation and they are integrated within separate components. These groups are connected mainly by the policy areas they occupy. The largest of these components belongs to the evangelical church. There we have, for example, actors such as the Polish Evangelical Church, Pentecostals and the Ecumenical Council. The largest group of actors revealing homophily in the policy areas are patient advocacy and health organisations. These include intellectual disability organisations, breast cancer associations, and deaf associations (the Polish Association of the Deaf), associations which advocate for people with diabetes, and many more. However, this may mean that our sample did not include the organisations through which the organisations shown in the graph are connected, and therefore due to this missing data we should treat the conclusions here with caution.

In other small components of the Polish network, we also find organisations promoting sports and ethnic minority organisations grouped around the same policy areas. The largest component of sports organisations includes four actors—the National Federation of Sports for All, the Society for the Promotion of Physical Culture, the Polish Association of Athletics, and the School Sports Association. Within the ethnic minority organisations we find the Union of Tatars in Poland—a Muslim minority and several German minority organisations.

Sweden and the UK

In this part of our analysis including the cases with relatively better-connected networks (‘giant components’, a connected component of a network that includes a significant proportion of the entire nodes in the network) of CSOs through interlocking leaders, we address the following question: Who are the most powerful actors based on their network position? Here we make comparative commentaries on Swedish and British cases instead of presenting the networks separately (see Table 10.2).

Table 10.2 Giant components in networks of leader interlocks in the UK and Sweden

When it comes to the global network level, the British network is a bit denser, meaning that there are more connections between the organisations via leader interlocks compared to Sweden. Also, the average degree (the average number of connections each node in a network has) is twice as high in the UK. Average distance tells us the average path between every pair of nodes in the network and nodes are slightly farther from each other in Sweden. While the British network is more strongly centralised in terms of degree, the Swedish network has stronger Betweenness centralisation, because paths between the nodes are longer, meaning that there are organisations occupying stronger betweenness positions. This can be also seen in individual measures below.

Regarding the eigenvector centralisation, it should be explained that Eigenvector centralisation is high when positions with high-degree centrality are connected to each other. Eigenvector centrality is increased by connections to high-degree neighbours, so when high-degree nodes are preferentially directly connected to one another, and low-degree nodes are preferentially connected to one another, eigenvector centralisation will be high. In other words, increases in assortativity—a preference for a network’s nodes to attach to others that are similar—are reflected in increases in eigenvector centralisation. This type of centralisation is higher in Sweden, because there are more such nodes that link to other nodes with the same number of links. Compared to the pure ‘star’ network, the degree of concentration in the Swedish sector is 66% of the maximum possible. This means that Swedish organisations are more than British ones concentrated around a few actors, who are the center around which other actors are concentrated.

Comparing the K-scores, we see that they are much stronger in the UK than the Swedish ones. The strongest red cluster contains actors that have twelve connections each, followed by the orange cluster containing eleven connections and then the yellow with nine connections (Fig. 10.3). However, the network in the UK has weaker assortativity than the Swedish network. This means that although there are strong K-cores they tend to connect to weaker nodes. So, in the English case there are strong elites within elites, but they do not claim exclusivity for their elitism, or at least to a lesser extent than the Swedish network.

Fig. 10.3

Network of interlocking leaders in UK. (Source: Authors’ own analysis)

In the Swedish case we have much weaker K-cores, the strongest links are three—red, two—orange and one link—blue (Fig. 10.4). But nevertheless, this network has a stronger tendency for elites of elites.

Fig. 10.4

Network of interlocking leaders in Sweden. (Source: Authors’ own analysis)

We now turn to the analysis of characteristics at the node level using individual measures such as betweenness centrality, degree centrality and eigenvalue. These measures have been normalised so that they can be compared across the two country contexts (Table 10.3). For example, the strongest Swedish actor in Betweenness is almost twice as strong as the English actor, etc. Of course, the group of actors distinguished here are also identified in the K-core measures.

Table 10.3 Organisations occupying central network positions in the UK and Sweden

In other words, these are the organisations in Sweden and in the UK who occupy the most powerful network positions. There is relatively high variability in betweenness centralities among organisations occupying the strongest network positions in Sweden (standard deviation 7.09) around the mean (27.3) in relation to the British sector (mean 17.6, standard deviation 4,1). This suggests that, overall, there are great inequalities in actor centrality or power, when measured in this way. The same applies to the eigenvalue measure (Sweden: mean 46.6, st. dev. 12,3, Britain: mean 34.9, st. dev. 7.2) but not the degree centrality measure which is almost the same within the most central actors in both countries (Sweden: mean 0.07, st. dev. 0.012, Britain: mean 0.11, st. dev. 0.016). Among elites of the elites, betweenness and eigenvector significantly differentiate the two countries studied here.

The Swedish side is more strongly varied in terms of organisational strength but also has stronger organisational actors at the top overall. However, this does not apply to the measure of degree, where there is no inequality between major organisations in our countries, but British organisations have, on average, more connections to other organisations in the network.

Looking at the identified organisations in the UK, using the three measures in Table 10.3 there is a consistent overlap between them. One important group includes the charities active within disability and health care (Blind Veterans, King’s Fund, St’Johns Ambulance and International Red Cross). We find also a group of charities involved in environment issues (Tusk trust, Wildlife Trusts and World Wildlife Fund). Other actors work with historical conservation (Societies of antiquities), culture (National Youth Theatre) and sports (Royal Yachting Association). Interesting to notice is that what these actors have in common is to have members of the royal family in their board as patrons or other honorary positions (e.g., honorary vice president).

Looking at the identified organisations in the Swedish case in the same table, we find a large number of umbrella organisations organising and representing CSOs across policy areas (Forum—ideburna organisationer med social inriktning and CIVOS) or as employers (KFO and Arbetsgivaralliansen). Another important group of actors is engaged in adult education with the umbrella organisation Studieförbunden and some of its members (ABF—Arbetarnas bildningsförbund, Sensus). Another relevant group is active in sports with the umbrella organisation Riksidrottsförbundet and its partner working with training and capacity building (SISU) and some of its members (Ishockeyförbundet and Gymnastikförbundet). Two other actors are involved in international development (We Effect) and consumers’ rights (Sveriges konsumenter). Also here we find an interesting pattern that these organisations tend to be linked by membership in each other.

Conclusions

In this chapter we have analysed the networks emerging from interlocking leaders among the most resourceful, national-level CSOs in four European countries. The analysis focused on the observed organisational links via interlocking leaders and we have identified primarily those related to communities of tightly connected organisations and the organisations that seem to bridge different communities of CSOs. The analysis included different national contexts, aiming to draw comparative insights informed by contextual knowledge.

For Italy and Poland, we found fragmented networks comprising of smaller components. While in Italy policy areas and ideological affinities between organisations explain the links between CSOs, in Poland we only find policy areas as a possible mechanism behind the organisational links. Contrast to the cases of Italy and Poland, in Sweden and in the UK we found giant components connecting a significant number of CSOs in each context. For Sweden we find that it is through the membership relations the leader interlocks occur, which means that the actors occupying the most powerful positions are the ones that have interest representation for the civil society sector either as a specific type of employers or as an agent for interest aggregation and political participation of CSOs at the national level for various issues. This interpretation rhymes well with the corporatist tradition of Swedish interest representation and advocacy culture (Arvidson et al., 2018). For the UK case we find that the network of interlocking leaders is upheld mainly by a particular type of leaders in this context, namely the ‘ambassadors’ and ‘patrons’ who have a rather symbolic function in the leadership of organisations and with a wider reach to multiple organisations, compared to the leaders with executive or representative roles. The organisations that occupy the central network positions in the UK are thus connected by these symbolic leaders with frequent linkages to the British royal family, an important source of power for accessing funding and other opportunities despite certain reluctance and ambivalence expressed around it by civil society leaders (Johansson & Ivanovska Hadjievska, 2022).

The case of the UK opens up of course for an understanding of linkages more related to the individuals than the organisations. Although our analysis has been mostly focused on the organisations, the symbolic role of the patrons and ambassadors in the UK seems almost self-evident due to membership in a high number of boards in combination with royal titles.

Based on our analysis, the civil society field in the UK and Sweden seems to be more consolidated and integrated via interlocking leaders compared to the Italian and Polish civil society with more fragmented characters. However, one important limitation of our study is that the organisations are selected using a set of criteria in identifying the most resourceful CSOs in each context, and therefore we do not capture potentially existing leader interlocks between the organisations included in our analyses with smaller organisations or organisations active at local/regional level or those with fewer resources. The fragmented characteristics of the Italian and Polish networks compared to the Swedish and British cases, which tend to mirror that fewer organisations are interlocked by leaders with multiple affiliations, might be understood as a consequence of other practices being employed for connecting organisations in civil society. In the Italian case, for instance, the presence of one large umbrella organisation for the whole civil society sector recognised both in civil society and by the state (Santilli & Scaramuzzino, 2021), might reduce the need of creating networks through cross-representation of leaders. These types of ties are not captured by our research method, whereas our method allowed us to identify the mechanisms of leader interlocks among the organisations in the Swedish and British cases.