Keywords

Introduction

Addressing the composition of elites is a task common to many elite studies (Hoffmann-Lange, 2018). This approach is about identifying who the elites are and what characterises them. In this chapter, we study the composition of civil society elites in four national and one supranational context in Europe, using a novel comparative dataset based on surveys of the top-level leaders of the most resource-rich civil society organisations (hereafter CSOs).

This study draws on a particular strand of research in elite studies focusing on the ‘elite-masses gap’. The basic assumption is that elites are different from the rest whether formulated in terms of ‘elite-masses gap’ (Müller et al., 2012), ‘elite-citizen gap’ (Dellmuth et al., 2022), or ‘elite-public gap’ (Kertzer, 2020). To study this gap, it is common to compare the composition and/or attitudes of an elite group with that of the general population (e.g., Dellmuth et al., 2022). A large gap between representative political elites and citizens has been suggested to potentially challenge well-functioning democracies. Best and Vogel (2018) link this gap to ‘political professionalisation’, which can be understood as a process at both the individual level and the structural level. While at the individual level representatives tend to become part of an occupational group, at the structural level the same group becomes established by formal and informal rules that define the group’s boundaries and collective identity. The creation of a ‘closed group’ of representative elites, for instance, makes some groups of citizens more likely to be selected into this elite group than others. In this sense professionalisation increases the autonomy of representative elites compared to their constituencies, thus making them less accountable (ibid.).

While the issue of the elite-masses gap has been particularly studied regarding political elites (Kertzer, 2020), it is also relevant for studies of the composition of civil society leaders. CSOs and their leaders are supposed to play a crucial role in bringing the voices of diverse groups into the political arena for a well-functioning, pluralistic democracy (Guo & Musso, 2007; Kohler-Koch, 2010; Smismans, 2003). Many CSOs claim, in fact, to represent the interests of specific social groups and to speak on their behalf. They seldom aim to represent the interests of the elites, but rather those of specific interest groups such as people with certain types of disabilities, retirees, sexual minorities, tenants, the homeless, and so on. Many others claim to be advocates for certain social causes, such as traditional values, social justice, gender equality, environmental protection, children’s rights, and so on. In both cases, CSOs and their spokespersons claim representativeness as a way to strengthen their leverage in communicating core messages in the public arena and in pursuing their missions, claiming to act as ‘transmission belts’ between citizens and policymakers (Albareda, 2018; Halpin, 2010). Although the leaders of CSOs are not necessarily appointed through formal elections by the masses in the same way as elected politicians, CSOs strive to establish ways of ensuring representativity of their leaders vis-à-vis their constituencies (Johansson & Lee, 2014).

Besides representativity based on internal democratic elections, claims of representation based on having a leader with shared characteristics and identities as those they represent can be justified in the representation of marginalised social groups. A descriptively representative leader can potentially bring the perspectives and experiences of disadvantaged groups directly into the political arena (Pitkin, 1967). In a similar way, Saward (2010) includes ‘mirroring’ as one sub-type of representative claims where a representative appeals to the similarity between the claimant and the constituency he or she claims to speak or stand for. An illustrative example of such an effort to achieve descriptive representation might be found in a federation type of organisation for people with disabilities, where the organisation makes sure that its executive board consists of persons with different types of disabilities (Johansson & Lee, 2014).

While we expect that CSOs will strive for diversity and inclusiveness in the social composition of their leaders, we expect the leadership positions of the most resource-rich CSOs in each national context to be mostly occupied by people belonging to the social categories that are often found in positions of power in society.

Based on these debates, we argue for the relevance of addressing the elite-masses gap concerning civil society elites. A key empirical question here is about the composition of civil society elites and how well they reflect the characteristics of and attitudes held by the general population. We thus study the civil society elite-masses gap as a representation of the differences between civil society elites and the general population in terms of socio-demographic characteristics and attitudes in four national contexts (i.e., Italy, Poland, Sweden, and the UK) and at the EU level. The research questions are as follows:

  • What similarities and differences do we find in the socio-demographic composition and attitudes of civil society elites in relation to the general population in each context?

  • How can we understand the elite-masses gap across different national contexts characterised by different roles of civil society and different degrees of professionalisation and at the EU level?

These questions are of key importance for civil society research. In fact, as with representative political elites, trends of professionalisation and long chains of representation might have contributed to creating a closed group of civil society elites. In line with Michels’ (2001) theory of the ‘iron law of oligarchy’, concerns have been raised as to the representativeness of leaders of major CSOs, thus problematising the possibly increasing distance between the leadership strata of civil society and their constituencies (Skocpol, 2003). The emergence of the skilled, professionalised civil society leaders, possibly resembling other political elites in their socio-economic disposition and career paths, raises the question of the capability of civil society leaders to actually deliver on the promises of democratic and pluralistic representation for their causes and constituents. In Michels’ (2001) view, the professionalisation of leaders would involve civil society elites becoming increasingly conservative towards societal change when it comes to their attitudes.

Method

Empirically, we analyse the composition of civil society leaders using cross-national survey data from the Civil Society Elite Survey (see the Appendix to this volume by Scaramuzzino and Lee) carried out within the research programme ‘Civil Society Elites? Comparing Elite Composition, Reproduction, Integration and Contestation in European Civil Societies’, funded by Riksbankens jubileumsfond 2018–2023. The respondents (N = 897) are top-level governing and executive leaders of resource-rich CSOs in Italy, Poland, Sweden, and the UK and at the EU level using a set of common indicators measuring financial and political resources. Our contribution is to empirically determine the composition of the group of leaders as a de facto civil society elite who occupy the highest leadership positions in the most resource-rich CSOs in the five contexts.

We focus both on the socio-demographic characteristics of the leaders (see the Appendix to this chapter) and on their attitudes regarding four core sociopolitical issues. By mapping out the composition of civil society elites in relation to that of the general population using another set of comparative data (the European Social Survey, 2018, Round 9 data), we are able to explore the civil society elite-masses gap. The cross-country comparison allows us to address both general patterns in civil society elites regardless of contexts and different patterns that can be understood through the lens of civil society regime theory and levels of professionalisation, as will be discussed later in the chapter.

In our analysis of the composition of civil society elites across the five contexts, we consider two sets of variables that are connected to the leaders’ socio-demographic background and to the leaders’ attitudes (see Table 3.1).

Table 3.1 Variables used in the analysis of civil society elites’ composition
Full size table

The socio-demographic variables include age, gender, education, and country of birth and relate directly to the issue of homogeneity and to how inequality in society is reproduced based on specific categories (Tilly, 1999). They also relate to theories about resources and capital that different individuals control based on their social position, such as cultural and social capital (Swartz, 1997). Based on a common understanding of political elites, which dates back to the work of Pareto in the nineteenth century, we expect a certain level of homogeneity among our leaders when it comes to socio-demographic characteristics (Best & Higley, 2018) and attitudes (Best et al., 2012; Gulbrandsen, 2019; López et al., 2020).

The attitude variables focus on the leaders’ attitudes concerning issues of democracy and equality. Among these variables, we consider social trust, ideological position, satisfaction with democracy, and gay and lesbians’ rights. The variables have been chosen partly because they allow us to compare the attitudes held by civil society elites with the general population in each national context through the European Social Survey (ninth round). Social trust relates to a classical outcome of civil society participation and engagement (Putnam, 2000). Ideological position, as a variable, places the respondent on a classical left-right continuum with a middle point. Satisfaction with democracy can be related to the classical function of civil society as a critical voice, watchdog, and counterweight to the state (Arvidson et al., 2018). Finally, gay and lesbian rights is an important dimension of human rights that creates polarisation in the public debate and that is not directly overlapping with an ideological positioning on a left-right scale.

The analysis is built around a bivariate analysis comparing the civil society elites with the general population in each national context. Measures of associations are presented for the correlation between the specific variable and national context of the civil society elite. As a measure of difference or distance between the civil society elite and the general population, we present a ‘gap-measure’ (cf. Müller et al., 2012) that we calculate as the difference between the mean value for the elite and the mean value for the general population.

Before we delve into the analysis of our data, it is relevant to briefly discuss the types of organisations that are represented in the sample that the leaders in the survey are leading. To the extent to which civil society leaders are expected to descriptively represent their constituencies, the type of organisations they lead will give some insights into what constituencies they are expected to mirror. Unfortunately, due to anonymity in our survey study, the responses of the leaders cannot be directly linked to a specific organisation or type of organisation. However, by looking at the sample of organisations in each of the national contexts, we can understand which types of organisations are represented among our respondents.

First of all, we find quite similar types of organisations and movements in all national contexts. Most of them work with general and broad issues such as health, social justice, poverty alleviation, the environment, human rights, international aid and development, sports, and so on. These organisations would have, in terms of the socio-demographic composition, broad constituencies involving diverse social categories. A very small number of organisations in our sample would relate to constituencies that are clearly defined in terms of socio-demographic characteristics, for example pensioners’ and youth organisations for age, women’s organisations for gender, and ethnic organisations for country of birth. The samples do not generally include any organisations that would have an over-representation of people with a higher education background (e.g., professional organisations). Hence, from a socio-demographic perspective, we would expect a large majority within the civil society elite in our study to represent constituencies that are diverse concerning our four variables of gender, age, country of birth, and education.

When it comes to attitudes, however, we would expect the leaders to embody the values and missions of the organisations they lead (and of their constituencies). The majority of the organisations in our sample have a progressive and solidaristic stance towards issues of social justice, equality, anti-discrimination, and human rights, although we find both organisations belonging to more traditional movements such as the Catholic organisations and organisations belonging to more secular and progressive movements (Santilli & Scaramuzzino, 2021).

Different Civil Society Contexts

Civil Society Regimes

One way of understanding different roles and ideals related to civil society across national contexts is through the concept of ‘civil society regimes’. The four countries considered in this study are often placed as examples of different civil society regimes, and their civil society sectors have different characteristics. 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 Italian civil society sector, as an example of a continental or corporatist regime, is mostly service oriented with a larger share of paid staff compared to Sweden. Likewise, the UK’s civil society is an example of an Anglo-Saxon or liberal regime and is characterised by the prominent role of civil society as service provider, with a larger proportion of paid staff than Italy and Sweden. Finally, Polish civil society, as an example of the Eastern or post-communist regime, is also service oriented, but with smaller workforce than the other countries and with a very small share of paid staff (Archambault, 2009; Salamon et al., 2017; Salamon & Sokolowski, 2018).

Based on this comparative contextual information, we argue that the civil society sectors in different countries can be understood as a segment of the professionalised labour market to varying degrees, which could entail more or less stratification and hierarchisation between the organisations with implications for the composition of the leadership strata. If we understand professionalisation of the sector as a precondition of elitisation, we expect that a civil society sector characterised by a larger employed workforce and a pronounced role as service provider—hence comprising a relatively well-defined labour market—is characterised by a leadership with a more pronounced elite status. From this point of view, we can expect the top-level leaders of the most resource-rich CSOs in Italy and the UK to be characterised by higher social positions and elite characteristics in their composition compared to Sweden and Poland. Italy and the UK would thus present a larger elite-masses gap compared to Sweden and Poland.

Conditions of Engagement of Civil Society Elites

We should, however, keep in mind that the regime-level characteristics from the existing literature relate to the civil society sector as a whole, while our survey respondents represent the most resource-rich organisations at the national level. Therefore, we complement the cross-contextual expectations based on the civil society regime literature with our first-hand survey data. Our data can provide us with relevant information about the conditions of engagement of the civil society leaders, which can be used to understand similarities and differences between the countries when it comes to the composition of civil society elites.

In Table 3.2, we present how the conditions of engagement among our respondents differ across our contexts in relation to a few variables that we operationalise as tokens of professionalisation. We look, for instance, at the extent to which the leaders receive economic compensation and if they support themselves mainly through their engagement in civil society. A larger share of leaders receiving economic compensation and supporting themselves with such compensation would indicate a more professionalised civil society sector. We also want to see how long they have been in the position of leadership and how long they have been engaged in the organisation. Shorter time in the position of leadership and shorter periods of engagement in the organisations are interpreted as tokens of professionalisation, in a sense that leaders are hired based on their competences and merits rather than their long-time commitment to the organisation. We also look at the share of executive leaders (e.g., secretary generals, directors, and CEOs) compared to representative leaders (e.g., spokespersons, chairpersons, and presidents), and a larger share of the former category of leaders is seen as a sign of a more professionalised civil society sector.

Table 3.2 Conditions of engagement of the civil society elite by national context
Full size table

The economic compensation of the leaders shows a higher level of professionalisation in Sweden and the UK than in the other contexts. Sweden has a higher share of leaders who receive economic compensation than the UK (75%), but the UK has a higher share of leaders who support themselves through their engagement in civil society (54%). This suggests that a larger share of Swedish leaders receive economic compensation of more modest amounts, often in the form of an honorarium rather than a salary. The UK is the only context in which a majority of the leaders support themselves through their civil society engagement. Leaders at the EU level are placed more in the middle on a continuum, while Poland and Italy have less professionalised leaderships with a minority of leaders supporting themselves through their engagement in civil society, specifically one in three for Poland and one in four for Italy.

We find a similar pattern when it comes to career trajectories with Sweden, the UK, and the EU presenting shorter trajectories both concerning how long the leaders tend to have held their positions and how long they have been engaged in the organisations they lead. The first measure suggests more frequent turnovers and possible labour market dynamics with individuals having occupied leadership positions for an average of five to six years in Sweden, the UK, and the EU. Italian and Polish leaders have on average been in leadership position for nine years. We find a similar pattern when it comes to the internal career trajectories in particular for the UK and the EU, with leaders having been involved in the organisations for 11 years, followed by Sweden with 14 years. For Italy and Poland, we find longer internal trajectories with Polish leaders having been involved for 17 years and Italian leaders for 21 years on average. Shorter internal trajectories can once again be understood as tokens of professionalisation and labour market dynamics, while longer internal trajectories can be understood as the opposite in the sense that qualifications related to long-term commitment to the organisation and the cause play a more important role. Finally, the share of directors among the respondents across the contexts points to similar comparative conclusions, with a greater share of executive leaders in the UK, the EU, and Sweden on the one hand and a smaller share for Italy and Poland on the other.

In conclusion, looking at the conditions of engagement among the civil society leaders who participated in our study we would expect more elite status of leaders and hence a larger elite-masses gap in the UK, the EU, and Sweden and less so in Poland and Italy. This expectation is considered in the following section where we analyse the socio-demographic composition of civil society elites in relation to the general population.

Socio-Demographic Background of Civil Society Elites

We address the socio-demographic background of the civil society leaders and the elite-masses gap in Table 3.3. First, we compare the mean age of our civil society leaders in relation to that of each context using European Social Survey data. We find significant differences between the civil society elites across the national contexts in this respect. The Italian leaders are the oldest on average (59 years old) followed by Sweden and the UK (57 years), the EU (53 years), and Poland (51 years). These differences between the mean values are statistically significant.

Table 3.3 Comparison of civil society elites and the general population across contexts: Socio-demographic background
Full size table

If we look at the average age in the general population in each context, however, we get a slightly different picture. Here it is important to remember that the European Social Survey includes only people aged 15 and older. The Swedish population is the youngest of all the contexts, and thus it is in Sweden that we find the largest age gap (12 years) between the civil society elite and the population, followed by the UK (10 years), Italy (9 years), the EU (5 years), and Poland (4 years).

The age span can be quite large, and thus it is relevant to also look at the age distribution to see to what extent the leaders’ ages tend to concentrate around the mean or if they are more dispersed. As it usually takes time to get access to elite positions, we expect the standard deviation of the age among our respondents to be smaller than in the general population, which is confirmed by our analysis of the data. The smallest variation we find is in the UK followed by Italy, the EU, Sweden, and Poland.

The second socio-demographic variable we consider is gender. When it comes to gender distribution, there are stark differences across the contexts. While it is only in Sweden that we find a higher share of female leaders among our respondents (58%), Poland is in second place with an almost equal representation of male and female leaders among the respondents (49% and 51%, respectively). Both sexes are also rather equally represented in the EU context (44% female and 56% male). In the UK and Italy, female leaders are the minority among our respondents—40% of our British respondents are female, while only 28% are female in Italy.

Because the share of the male population is very similar across the national contexts, variations between national contexts when it comes to the share of male leaders is a direct indication of different gender gaps across the contexts. We measure the gender gap as the difference between share of males in the general population and the share of males among our respondents. Only one country context, Sweden, presents a negative gender gap with a 9% lower share of males in the civil society elite than in the general population. We find a small gender gap in Poland with only 3% fewer males in the civil society elite than in the general population. We find larger gender gaps at the EU level (6%) and in the UK (11%) and especially in Italy (24).

Studies of elites have found extensive evidence for the importance of educational capital, and in some contexts even attending specific higher education institutions, for becoming part of the societal elite (e.g., Ellersgaard et al., 2013). It turns out that the absolute majority of leaders of resource-rich CSOs in all of our contexts have had higher education. Poland and the EU level stand out to some extent, with over 90% of all respondents having had higher education. Next follows the UK (88%), Italy (79%), and Sweden (77%). The lowest share of leaders with a higher education background found in Sweden might be explained by the democratic ideal upheld in the popular movement tradition characterising Swedish civil society (Lundström & Svedberg, 2003). Compared to the general population, the education gap is larger in Italy than the UK due to the smaller share of highly educated people in society at large in Italy.

The last socio-economic variable is place of birth, and we look at the share of leaders that are native born in each national context. We find the smallest share of native-born leaders at the EU level with 80% of leaders born within the EU, a figure that is 8% lower than the average in the EU’s general population. This might be explained by the large number of international organisations present at the EU level. Poland also has a negative foreign-born gap (−1%), but the numbers are very small both in terms of share in the civil society elites (98% native-born leaders) and in the general population, so the pattern must be taken cautiously. The UK has the lowest share of native-born leaders (87%) among the national contexts and also a relatively small gap of only 4%. The same goes for Italy due to the larger share of native-born persons in the general population. In Sweden, we find a relatively small share of leaders born in Sweden (91%), but considering a relatively higher share of foreign-born persons in the general population we find the largest gap in Sweden.

In sum, we find that each of the socio-demographic characteristics of the civil society leaders we looked into has its own pattern when it comes to the differences across the national contexts and in relation to the general population. A higher gap concerning our variables suggests that the civil society elite can be characterised as ‘exclusive’, in the sense that there is an over-representation of dominant groups in society (older, male, highly educated, and native-born leaders), while a lower gap might imply a more ‘inclusive’ national context in terms of the composition of civil society elites. The results show no uniformly cross-cutting pattern for any of the variables considered. Sweden stands out as being the most inclusive when it comes to gender and education but the most exclusive when it comes to age and country of birth. Also, Poland and the EU level have mixed results showing an inclusive tendency in terms of age but an exclusive tendency in terms of educational background. A pattern common to all national contexts is that the civil society elite tends on average to be older than the general population and that they are to a larger extent highly educated. When it comes to gender and country of birth, we find different patterns in the national contexts.

Attitudes of Civil Society Elites

We address the attitudes of the civil society leaders and the elite-masses gap in Table 3.4. The first variable that we focus on is social trust. A general pattern found across contexts is that the level of social trust among the civil society elites follows the national pattern, and the civil society elites in low-trust countries, for example Italy and Poland, have lower trust than the civil society elite in high-trust countries like the UK and Sweden. It is also clear in all national contexts that the civil society elites have higher social trust than the general population and that the gap is larger in low-trust countries.

Table 3.4 Comparison of civil society elites (CSE) and the general population (POP) across contexts: Attitudes
Full size table

The second variable we focus on is ideological position on a left–right scale represented by numbers ranging from extreme left (0) to extreme right (10) with a middle point (5). We find that civil society elites in all contexts position themselves more to the left compared to the general population. In addition to this commonality, we find differences between the contexts with the Italian leaders being more to the left in terms of a larger gap, followed by the EU level, the UK, Poland, and Sweden, whose civil society elites are comparatively closer to the middle point of the ideological scale.

With regards to the variable measuring the extent to which respondents are satisfied with the ways in which democracy functions in one’s country, we find a consistent pattern across the contexts where civil society elites are less satisfied than the general population (except for Italy where the mean value is the same). The most satisfied with how democracy works in their country are the Swedish leaders, followed by Italian, EU-based, British, and finally Polish leaders, who are least satisfied. It is also clear that the gap in relation to the general population is not that large except for in Poland where civil society leaders are on average 2.1 points less satisfied on a scale from 0 to 10.

The last variable we compare is the extent to which respondents endorse equal rights for the gay and lesbian population, and we find that in all national contexts the civil society elites are more progressive on average than the general population. Following the national pattern, we find slightly more conservative views among our leaders in Italy and Poland than in the other national contexts. The largest gap we find is in Poland where we observe the most conservative attitude towards this issue. In fact, Polish civil society elites are on average 1.1 points more progressive than the general population on a scale from 1 to 5.

When it comes to attitudes regarding key issues of democracy and equality, we thus find a much clearer pattern with Poland being the national context with the largest gap between the civil society elites and the general population (except for ideological position) and Sweden being the context with the smallest gap. Furthermore, we consistently find that the civil society elites tend to have higher social trust, are more leftist, are less satisfied with democracy, and are more progressive towards gay and lesbians’ rights than the general population in all national contexts (except for satisfaction with democracy in Italy).

Conclusion

In this chapter, we set out to explore the composition of civil society elites across different countries and at the EU level in comparison with the general population in each context in order to address the civil society elite-masses gap. We depart from a theoretical perspective where the civil society elites are expected to be representative of the general population in terms of their social composition and attitudes. At the same time, we expect the empirical results of our analyses to be more complex, informed by the perspective of a possible tendency of elitisation of civil society leaders driven by professionalisation of the top leaders in the sector. In this concluding section, we discuss the main results of our analysis of the composition of civil society.

When it comes to the socio-demographic backgrounds of the civil society leaders, we find both inclusive and exclusive patterns in three of the contexts, that is, in Poland and Sweden and at the EU level. This shows that civil society elites can be characterised as both inclusive and exclusive social groups when it comes to different dimensions and different contexts. With respect to age, civil society elites in all contexts are of older age, and this can be related to the long career trajectories that are required for people to reach the leadership positions of the most resource-rich national-level CSOs. The fact that civil society elites in all contexts have partaken in higher education to a much greater extent than the general population can be related to a demand for certain knowledge that is best acquired through academic degrees.

Two of the country contexts present exclusive patterns concerning all of the socio-demographic background variables among the civil society elites, that is, Italy and the UK. The leadership strata in these contexts are characterised by an over-representation of male, older, highly educated, and native-born leaders. These exclusive patterns are partly in line with our expectations based on these countries’ civil societies being more service oriented and professionalised to a higher degree compared to Sweden and Poland. When it comes to the UK, it is also in line with our own survey data regarding the conditions of engagement of civil society leaders.

We find a clearer pattern across countries when it comes to the civil society elite’s attitudes and how they fare in relation to those of the general populations. A consistently higher level of social trust among our leaders compared to the general population can be interpreted from the perspective of a widely understood role of civil society as the fabric of social trust in society (Putnam, 2000). The more leftist positions of the leaders compared to the general population can be related to the historical development of many CSOs, originating from progressive social movements engaging for social justice and the emancipation of different minority groups (Ruzza, 2020). Dissatisfaction with democracy among the leaders can be understood in relation to the critical role that is ascribed to civil society actors in liberal democracies. It could also be seen in light of the neoliberal turn of the last decades in many countries’ social policies that has negatively and disproportionately affected the weaker social groups for which many CSOs advocate (Chancel et al., 2022; Piketty, 2014). Increased inequalities both globally and within countries have been reported over the last 40 years (Chancel et al., 2022). Finally, progressive attitudes towards gay and lesbian rights can be related to the historical legacy and focus on human rights and the fight against discrimination as discussed above.

An elitisation of civil society leaders would, in accordance with the elite literature, suggest that the elite-masses gap is more in the direction of the elites being more conservative and satisfied with the status quo than the masses. In fact, this is the thesis brought forward by Michels (2001) as part of the ‘iron law of oligarchy’. Compared to the general population, our data tend to contradict such expectation, suggesting instead that civil society elites have more progressive attitudes and are less satisfied with democracy than the general population. It would of course be interesting to see to what extent the volunteers and activists within CSOs would have even more ‘radical’ attitudes.

Comparing our results concerning the two sets of variables in our study, we thus find a certain homogeneity across countries when it comes to the civil society leaders’ attitudes rather than their socio-demographic backgrounds. Trying to empirically explain this result would require another round of study, but there are some possible explanations. One hypothesis would be that the composition in terms of socio-demographic background is partly a product of exogenous factors, for instance gender and ethnic-based relations in each national context. The large share of women at the top of Swedish civil society could be a consequence of both exogenous factors such as the stronger position of women in society as well as endogenous factors such as the focus on voluntary engagement rather than professional engagement in the civil society sector. Following a similar logic, Italy would be the opposite case with a low share of women at the top due to a generally weaker position of women in the labour market and in positions of power as well as a more professionalised, service-oriented civil society sector.

The composition in terms of attitudes and the more homogeneous pattern that our results show in this respect could be interpreted as a consequence of endogenous factors within the civil society sector, such as the historical legacy and ideological profile of the organisations. Here, two separate and not mutually irreconcilable hypotheses could be put forward. On the one hand, it is possible that holding a certain set of attitudes is an important selection criterion for becoming a leader, creating mechanisms of ‘ideological control’ of the people who are appointed, elected, or recruited to higher positions in the organisations. These mechanisms would function in a similar way as other requirements such as higher education, thus creating more or less formal selection criteria for career advancement (cf. Johansson et al., 2022). On the other hand, it is also possible that a mechanism of socialisation as well as long internal career trajectories (between 5 and 12 years before becoming a leader on average) contribute to shaping the relatively homogenous attitudes among the civil society leaders. Adhering to these attitudes could become part of a collective identity that is also part of a professionalisation process, as described in the introduction to this chapter.

The similarities in terms of attitudes considered in our study cut across different civil society regimes as well as the different degrees to which civil society sector is professionalised. This suggests that there exists a possible core value community of the civil society elites beyond specificities of different national contexts that shapes the composition of the top-level civil society leaders. What might be called a ‘civil society ethos’ could possibly distinguish civil society elites from other elites.