Trajectories of Civil Society Leaders in Italy: Individual Careers, Organizational Structures, and Ideological Affiliations

The aim of this paper is to explore the relationship between individual leaders’ career trajectories and organizational structure and identity in Italian civil society organizations (CSOs) active at national level. The analysis in this qualitative study draws on two sets of data, semi-structured interviews with leaders on their career trajectories and their understanding of what it takes to become a leader of a CSO and short biographies of leaders’ career trajectories. Three main representational logics are idenitified: within: multi-level, within: member-organization, and outside: supporting organization. The two first logics are based on a trajectory within the organizations either departing from the local and regional levels of the same organizations or within one or more member organizations. The third logic is based on a trajectory outside the organization that is marked by the ideological affiliation of the president through leading positions in other organizations within the same movement or field.


Introduction
The aim of this paper is to explore the relationship between individual leaders' career trajectories and organizational structure and identity in Italian civil society organizations (CSOs) active at national level. The analysis in this qualitative study draws on two sets of data, semi-structured interviews with leaders on their trajectories and their understanding of what it takes to become a leader of a CSO at national level and short biographies of leaders' previous experiences and careers. The focus in the analysis is on the presidents' career trajectories in terms of previous experiences of engagement in civil society and in other spheres of society, that is, the public and business sectors and on the CSOs they lead, their organizational structure, identity, and historical development.
Studies of civil society leaders' career trajectories are few. While some studies from the US have focused on the career trajectories of executive leaders in the non-profit sector (Norris-Tirrell et al., 2018;Suarez, 2010), studies of presidents' or chairpersons' career trajectories are very few (see Lindellee & Scaramuzzino, 2020 for an example). This might be a consequence of the fact that the presidents are often understood as voluntary or unpaid leaders compared to the employed, paid executive leaders (cf. Kreutzer & Jäger, 2011;Meeuwisse & Scaramuzzino, forthcoming and the concept of career is often related to professionalism rather than voluntarism. Nonetheless, we find some relevant studies looking at the qualifications of presidents and chairpersons. Some of these studies focus on the recruitment of executive managers and/or board members, including the chairpersons (Abzug & Galaskiewicz, 2001;Iecovich, 2005) while other studies look directly at the leaders' characteristics (Iecovich & Bar-Mor, 2007;O'Regan & Oster, 2005;Prouteau & Tabaries, 2010). The only Italian study on the profiles of civil society leaders was based on interviews with presidents, vice-presidents, and board members (Colozzi & Prandini, 2008). Most of these studies of civil society leaders careers, characteristics and recruitment however, tend to not relate the leaders' identities, biographies, and orientations to the organizations they lead and represent (Orlandini, 2010).
The contribution of our study to previous research is hence twofold. Firstly, that we explore the career trajectories of CSO presidents instead of the ones of executive managers. Although research has pointed at the increasing professionalization of boards (Abzug & Galaskiewicz, 2001;Guo & Musso, 2007;Suarez, 2010) we expect presidents' career trajectories to be at least in part shaped by the organizations' organizational structure and identity. This hypothesis is motivated by CSOs being expected to be guided by religious, political, or cultural values, and their leaders to represent these values through their history and personal commitment (cf. Anheier, 2005). Presidents in particular are the organizations' spokespersons and have a political and representative role. Their personal charisma, skills, networks, and competences are assets that the organizations can make use of, but they need to be in line with their organizations' identities, goals, and missions (see XXXX, forthcoming).
The dynamic relationship between individual representatives and organizations has been conceptualized as leaders being "organisational centaurs" having the body of the organization and the head of a person (Ahrne & Papakostas, 2002). This metaphor allows us to understand the complex relation between the career trajectories of the leaders representing the organizations and the organizations' structure and identity. This brings us to our second contribution, that we explore the relationship between individual presidents' career trajectories and organizational structure and identity.
The choice of Italy as national context for the study has reasons but also implications. Previous research has shown that Italian civil society has always been shaped by cultural and ideological cleavages that polarize it (cf. Bassoli, 2017;Bassoli & Theiss, 2014;Tosi & Vitale, 2009). The polarization has mainly been studied at local level (Bassoli, 2017;Bassoli & Theiss, 2014). Some scholars have shown that at national level it is easier to find what Tosi and Vitale (2009) call cultural métissage (or contamination); that is the fact that "different families of political actors have strong cross-cultural relationship" (Bassoli, 2013, p. 11). However, the extent to which cultural affiliations still play a central role for the appointment of leaders at national level, is still debated.
Italian civil society, often described as third sector-The other two sectors are the public sector and the business sector-has been characterized the last 20 years by increasing dependence on public authorities and a process of professionalization and institutionalization. This development is particularly evident in the sector of social cooperatives (Ascoli et al., 2002;Borzaga & Fazzi, 2011), and the number of employees in the Italian third sector increased by 61% between 2001 and 2011 (Barbetta et al., 2016). Another trend in the Italian third sector relates to the traditional "collateralism" between the political parties and important segments of civil society (Della Porta, 1995Diani, 1996). On the one hand, the 1990s can be seen as a landmark of change of Italian civil society because of the collapse of the party system, which weakened the institutionalized relation between civil society and political parties and brought the development of a more autonomous civil society. On the other hand, it has also been stressed that the Catholic and the post-communist ideological affiliations of CSOs are still very much relevant in Italy (Bassoli, 2017;Biorcio, 2007;Polizzi & Bassoli, 2011;Tosi & Vitale, 2009).
The tension between a service oriented and increasingly professionalized sector and a polarized and ideologically segmented civil society, makes Italy and interesting case for a study of the career trajectories of Italian civil society leaders, looking at previous professional experiences, engagement in civil society, and the extent to which the ideological affiliations of the organizations play a role in the development of these trajectories. We pursued this aim by analyzing the career trajectories of the presidents of 30 national organizations selected because they are among the organizations with the highest level of status and recognition (both in civil society and by the state) at national level. We understand these organizations to be in need of the skills and expertise that often drive professionalization in the civil society sector at the same time as their position at the top of the hierarchy of Italian organized civil society presupposes long career trajectories and the capacity to represent important sections of Italian organized civil society.
The next section presents some concepts that we will use in the analysis of the empirical data. The following section explores the history and development of Italian civil society and provides the general background we need to be able to understand the organizations' structures and identity. Then follows a description of the methodology used. Thereafter we present the analysis of the empirical data and our conclusions.

Theoretical Framework
To understand the career trajectories of the leaders reaching the apex of power in this sector, we need to address their own career paths in relation to the development of the organizations they lead and represent. In this section, we present the three analytical concepts that allow us to explore the relationship between individual leaders' trajectories and organizational structure and identity in Italian civil society, namely representation, organizational structure, and pillar.
Drawing on organizational theory, we acknowledge a tension between the individual and the organization they belong to and represent. This tension can take different forms in different organizational contexts and depending on the role held by the individual. The tension is often assumed to be weaker in non-profit organizations, where the work is often based on a voluntary effort and where the individuals tend to share the organizations' objectives and values (Ahrne, 1999). In this sense, there is a demand for value congruence between a CSO and the people who are supposed to lead it (cf. Anheier, 2005).
Presidents of CSOs have a representative role, which can be understood as to act ". . .in the interest of the represented, in a manner responsive to them" (Pitkin, 1967, 209). Representation can be claimed in many different ways (Saward, 2010), and one common strategy to all the organizations we are dealing with in this paper is to pursue their presidents' representativeness through election. Studies of CSO-leaders presentations at the time of appointment show that people are elected to leadership positions based on a claim that because of their background, values, capacities, and objectives they are able to act in the interest of the members and in a manner responsive to them (e.g., XXXX, forthcoming).
While CSOs tend to organize individuals based on common values, goals, interests, and identities in many countries, the sector tends to assume hierarchical structures by means of meta-organizations, that is, organizations that have other organizations as their members (Ahrne & Brunsson, 2005). Such organizations are often set up to allow better cooperation between organizations, to allow stronger influence in society or politics, and to achieve status and recognition (Ahrne & Brunsson, 2005).
All of these motives might be relevant for the Italian civil society sector, which has created its organizational structure mirroring the administrative multi-level system of the state (municipal, provincial, regional, national, and European). This might be interpreted as a way of being able to address public policy at all levels of public administration in a coordinated way. One example of this would be the area of international aid in which we find different federations of organizations, that is, a national meta-organization organizing CSOs from the Christian-democratic movement (FOCSIV) and a meta-organization organizing CSOs from the communist movement (ARCS). There is also a metaorganization (AOI) organizing all meta-organizations working with international aid regardless of ideological orientation. Finally we also have a so-called "fourth level" network, a permanent civil society platform representing the whole sector at the national level, namely the Forum del Terzo Settore (Ranci et al., 2009). This fourth-level network includes "networks of the second and third level." Networks of the third and fourth levels are umbrella organizations that have as their members lower-level networks of organizations, while second-level networks have as their members individual membership-based organizations. The Forum del Terzo Settore is hence at the top of the hierarchy of meta-organizations.
In this Russian-doll system of organizations, the issue of representation becomes troublesome. On the one hand, we have the issue of diversity of organizations as described by Ahrne and Brunsson (2005, p. 432) who claim that organizations are based on the idea that they should differ by specializing their tasks or competences. Hence, organizations tend to be more clearly differentiated than human beings who do not have a fundamental need to be perceived as different from one another. Representation within a meta-organization, which has other organizations as members, can become more troublesome than within individual membership-based organizations, especially if the member organizations are very different when it comes to values, norms, and areas of activity. The more complex the structure of a meta-organization is, for instance when it comes to levels of governance, policy areas, and ideological orientation, the longer the chains of representation become (cf. Johansson & Lee, 2014;Kröger, 2008), and this might challenge claims of representation by leaders. Accordingly, different types of meta-organizations need to handle the issue of representation differently following different representational logics.
Finally, because CSOs are value-based and ideologically driven, the leaders' trajectories can be affected by the CSOs' origins and ideological affiliations to different movements, for instance, the Christian-democratic and the communist movements in Italy. To understand these different movements and traditions, we adopt the concept of pillar to describe a set of organizations, institutions, and movements linked to the same cultural and ideological framework (cf. van Hooren & Becker, 2012). We argue that in a polarized civil society sector as the Italian one, their belonging to one of these pillars forms many organizations' collective identity (cf. Polletta & Jasper, 2001). The concept of pillar allows us to move from the organizational level to the societal level of cultural and ideological cleavages that have shaped civil society in Italy (cf. Bassoli, 2017;Bassoli & Theiss, 2014;Tosi & Vitale, 2009).
While these concepts will help us understand the organizations' structures, the representational roles of the leaders in the organizations, and the role of the cultural and ideological framework for the organizations' identity, we need to frame Italian CSOs more clearly in the historical context of their development.

Italian Civil Society Between Pillarization and Professionalization
Modern civil society in Italy has its roots in the 19th century during the formation of the modern Italian state. The Catholic Church has played an important role (Barbetta & Maggio, 2008). The Catholic institutions were service-oriented actors able to compensate for the absence of public welfare services and they maintained their power and influence after the formation of the new Italian state. For most of the 20th century Italian civil society has thus been dominated by the Catholic Church, but also by the cooperatives, trade unions, and mutual benefit societies set up by the workers' movement with a socialist ideological orientation, emerged mainly in the north and the center of the country at the beginning of this century (Barbetta & Maggio, 2008;Cartocci & Maconi, 2006).
The end of World War II and the fall of the fascist regime gave a pivotal role to political parties characterized by an imperfect two-party system with the Christian Democratic Party (DC) permanently in power as the "government elite" and the Italian Communist Party (PCI) permanently in the position of the "constitutional opposition" (Verzichelli, 2016). These two parties had an important role in the control of the whole society (Galli, 1966;Sartori, 1982) to the point that they were "the principal channels of access to the bureaucracy and the principal transmission belts in the allocation of resources from centre to periphery" (Newell, 2015, p. 9) and able "to penetrate the interstices of civil society and the state" (Newell, 2015).
These two parties supported and were supported by specific civil society areas, that is, the Catholic and the socialist organizations (Biorcio & Vitale, 2016;La Valle, 2006;Manoukian, 1968). These relations were so strong that they influenced the construction of the elites of the two parties and the elites of their respective civil society areas. Biorcio and Vitale (2016) show that these parties had different ways to produce leaders. The PCI trained its members internally thus producing the leaders of the related civil society actors. For the DC, in contrast, it was the Catholic civil society that trained and provided the political party with some of their best leaders (Biorcio & Vitale 2016). During the 1960s and 1970s, the interdependence between civil society and the political parties became however more complex and new associative actors were created.
From the 1970s and onward, this tight relationship between parties and civil society started however to loosen up and solidarity-based actors and their leaders became more independent from their political counterparts. The cultural, social, and political mobilizations of the 1960s and 1970s brought to public attention new social issues such as the rights of the most vulnerable groups. In that context, like in many other European countries, much service production was transferred to different types of solidarity organizations within civil society, namely the third sector. This period is in fact characterized by the emergence of new and diversified associations, mainly voluntary organizations but also social cooperatives, a type of organization not bound to a particular form of organization or governance (Borzaga & Fazzi, 2011. They often assumed a mixed organizational form of both workers and volunteers, and the majority of them were semi-professionalized organizations (Borzaga & Fazzi, 2011. The new organizations were mostly present in welfare services, in particular in health care, social services, and assistance to disadvantaged groups such as people with disabilities, people leaving in poverty, and the homeless (Borzaga & Fazzi, 2011). These actors had a high degree of autonomy in their strategies and orientations, and they were increasingly supported by public funding (Borzaga & Fazzi, 2011, p. 415).
With the collapse of the Italian post-war parties and of the whole First Republic in the early 1990s, new CSOs, more institutionalized than the previous ones, became prominent. What has been called the pioneering phase gave way to that of the expansion and recognition of the Italian third sector (Borzaga & Fazzi, 2011). In fact, while this period was characterized by a general and marked loss of trust in political parties, which is confirmed by the decreased membership in political parties (Mair & van Biezen, 2001), the number people who participated in civil society actions increased (Biorcio & Vitale, 2016;Istat, 2013).
During this period, the number of organizations and their diversification increased, mostly because of the low public coverage of social services. As in the past decades, the historical organizations previously in place linked to the Church and to the workers movements had a privileged relationship with public authorities, but during the 1990s other actors became increasingly prominent acquiring "a new ability to orient political choices, to shape agenda setting in the public sphere and to filter media communication" (Ruzza, 2010, p. 263). From the year 2000, the structure of civil society changed further. This last period is characterized by a continuation of the marked dependence of the third sector on public authorities and by a process of professionalization of the whole organized civil society, but in particular of the social cooperatives, which are the main local actors of the third sector (Ascoli et al., 2002;Fazzi, 2013).
This short overview of the historical development of Italian civil society presents a tension between a tradition of closeness between most organized civil society and the two political parties, forming the two pillars, and a development of diversification, dependency of public funding and professionalization of the third sector.

Data and Method
The analysis focuses on the relationship between individual leaders' career trajectories and organizational structure and identity in Italian civil society organizations active at national level. The study draws on a sampling of relevant organizations active at national level in Italy. The organizations were selected because of their high level of status and recognition both in civil society and by the state. In order to identify them we have made use of five indicators that allow us a broad and complex interpretation of status and recognition in civil society following the Multi-dimensional Measure of Resource Stratification in Civil society (MMRSC) (see XXXX, 2020). The first set of two indicators measured different forms of internal status and recognition: (1) Organizations that held posts in decisionmaking bodies within umbrella organizations in specific policy areas (so called "third level networks"); (2) organizations that were members in the umbrella organization representing the civil society sector (so called "fourth level network"). The last three indicators measured external status and recognition: (1) organizations that were included in the register called "5 × 1000" of the Italian tax agency, a scheme that gives to the Italian people the opportunity of devolving 0.5% of their taxes to organizations; (2) Organizations that were included in specific ministries' registries for consultations; (3) Organizations that held posts in the Council for the Third Sector, the consultation body between the state and civil society. For our study we selected the 45 CSOs that fulfilled at least three of our five indicators (from the total sample of 291 CSOs) (see XXXX, 2020) as the ones with the highest level of status and recognition (both in civil society and by the state) at national level, in accordance with our aim.
Previous research on the Italian civil society has stressed the high degree of informality of its sub-national governance showing that access to decision and powerful positions and arenas is regulated by complex, informal and opaque procedures to the point that some scholars talk about « governance Italian-style» (Polizzi & Bassoli, 2011). With that in mind, considering our focus on relationship between individual leaders' career trajectories and organizational structure and identity in Italian civil society, we believe that the qualitative research methodology best suits the analysis of the logics at the base of CSOs Italian leaders' trajectories.
The study is based on two sets of data: interviews and short bios. To get access to interviewees for the semi-structured interviews we contacted all 45 CSOs asking for interviews with the presidents. We were able to interview 15 presidents from this sample and by snowballing we were able to include three more presidents for organizations that fulfilled less than three indicators. The interviews focused on the previous professional experiences of the leaders and their previous engagement in civil society, the structure of the Italian third sector, the networks, alliances and conflicts inside and outside the third sector, the election processes, and upward mobility. We were also offered to interview four directors of the CSOs, which we accepted as to have the directors' perspective of their trajectories in comparison with that of the presidents (which is the focus of this study). Through snowballing, we also interviewed three more directors for organizations that fulfilled less than three indicators. In total 25 interviews were conducted with presidents and directors of Italian CSOs active at national level.
In order to gain another perspective of the leaders' career trajectories we also focus the analysis on the 30 presidents for which biographic information were public and available on the web during the period from January 2019 to April 2019. The sources for this analysis in the web were the organization's webpage, LinkedIn, Wikipedia, and any professional profile/CV from the leader's own webpage or other employers. The number of 30 short bios has been chosen as it is large enough to provide variation of organizations and career experiences and small enough as only including organizations with high levels of status and recognition. Furthermore, the relatively low number of biographies allows for an in-depth analysis not only of the career trajectories of the presidents but also of the characteristics and historical development of the organizations.
The interview data was analyzed qualitatively and categorized based on the leaders' different understandings of the importance of organizational structure and identity for the career trajectories of the leaders. The data was coded and analyzed employing an open coding. In order to understand the relationship between individual leaders' career trajectories and organizational structure and identity in Italian civil society, we developed a thematic analysis by grouping the codes into larger categories such us "organisational structure" and "belonging and pillars." In the discussions reflections on the own trajectories and on the more general dynamics in Italian civil society were often intertwined. To get a clearer understanding of the actual trajectories of the leaders we have chosen to triangulate this analysis with that of the short bios of the leaders. For the analysis of the biographical data, we used a qualitative approach looking at the individual trajectories in terms of education, experiences of engagement, employment, and leadership. In the analysis of this data we explored also the relationship between the individuals' trajectories and the organizations' structures and networks, development, and belonging to different pillars. We argue that each of this sets of data provide a relevant but also partial description of the phenomenon. In the interviews, the informants speak both from their point of view as leaders but also as people knowledgeable of the sector. They give us their understanding, from their point of view, of the relationship between individual leaders' trajectories and organizational structure and identity in Italian CSOs active at national level. The short bios can be seen as less biased, although they of course mirror the aim of the compiler, whether it is the leader him-/ herself or the organization. It is of course not known if experiences that are not mentioned in the bios are missing because they were not relevant to mention or because the leader actually lacks them. By triangulating with these two sets of data (including interviews with directors) we argue that we are able to develop a more comprehensive understanding of the phenomenon (cf. REF).
For ethical reasons we have in the analysis anonymized the leaders that have been interviewed while the leaders for whom we have collected short bios are presented by organizational affiliation.

The Interview Study
Following our aim, the interview data was analyzed using two overarching categories that allow us to understand the role played by the organizational structure and the ideological affiliation both in the career trajectories of CSOs Italian leaders and in the Italian third sector in general: "organizational structure" and "belonging and pillars." Most of the leaders (22 out of 25) interviewed presented their career trajectories as internal to an organization or to a "political culture." "A person invests in one organization and stays there. There is a choice at the beginning of the career. I moved from my organization xxx to xxx because my first organization became part of xxx. You choose to which organization you want to belong and this one will become your home, and basically remain, remain your home." Here we can see that the organization is considered as a home that you choose from the beginning of your engagement in civil society. There are also references to organizations being part of a wider cultural and political group. This aspect seems to differentiate the Italian context from the international context, as stressed by a leader currently engaged at EU level and previously president of two Italian major organizations: "The substantive issue is that in Italy mechanisms work by co-optation, adherences and belongings to certain groups of belonging and interest. There are also some battles between these groups but mechanisms in Italy are well structured and rather specific."

Belonging and Pillars
The terms "belonging" seems to be a key element for understanding the career trajectory of most leaders (19 out of 25) according to our interviewees, including their own trajectory. According to the words of these interviewees, upward mobility in the Italian third sector is based on belongings and on the recognition of both leaders and organizations. On the contrary, not having an affiliation represents for some leaders a problem for their career: Belonging to a cultural group or, using the words of this leader, to a "family" seems to become the condition sine qua non for being recognized. Before knowing you, people need to know from where you come and of which group you are part. However, for other leaders (3 out of 25) the issue of belonging to a specific cultural group is no relevant anymore for the career trajectory of the leaders: "this aspect was important until the 90s because the third sector was quite linked to political parties, whereas nowadays we are independent of them and of ideologies." This last quote suggests that the rupture between Italian civil society and political parties has led to a reduction of the relevance of ideological affiliation. Expressions such as "family" and "political culture" and "group" used by interviewees can all be linked to the concept of pillars, meaning a group of actors linked to the same cultural and ideological framework. All the leaders interviewed agree in saying that pillars marked the history of the Italian third sector. However, this issue is still controversial, as shown by the following quote: "Isn't true that in some organizations you have a balance between Catholic leaders and Secularist leaders in order to be able to work? maybe we'd better say that all -not all actuallyof us are working to reduce these differences and operate all together. We are not linked to political parties as in the past, but this aspect is still relevant. It is quite different compared to the past but still relevant. The issue is that we are dealing with a sort of internal information, an unsaid. We know but we don't talk about it." The last quote highlights that pillars' relevance may also depend on the structure of the organization as some organizations (including the one the interviewee above represents) have the specific function of overbridging ideological divides aiming at a broad representation of the sector for instance within a specific policy area such as international aid.
Organizational structure. Some organizations born in the 1990s as networks have the function of unifying the third sector beyond the ideological conflicts that marked this context. This kind of organizations can be seen both as a consequence of the end of the political polarization and as a factor contributing to it. Due to the overbridging function of these organizations, their leaders have the sometimes difficult task of managing the rests of this traditional divide in civil society.
The fact that the relevance of pillars is also linked to the structure of organizations is confirmed by other leaders (18 out of 25). According to another interviewee "the organizations with a long history and a hierarchical structure (with regional and local chapters) are more linked to political and cultural groups than the ones operating only at national level." This leader makes a distinction between organizations that have a national headquarter and then regional and local chapters throughout the whole country, and organizations that have only a national headquarter. According to him the first typology is characterized by internal leaders' trajectories claiming some form of representation of the local realities whereas the second one would be more open to external leaders, claiming expertise in managing funding and handling communication and media.
Another director confirms this assumption stating that "some organisations are more structured and therefore have more sedimented forms of selection of leaders based on internal logics, whereas others, such as volunteering associations, are more fluid in the selection and often choose their leaders from different organisations" International organizations, which are mainly active at national level and don't have local chapters, also seem to follow dynamics not based on pillars: "In international organisations you don't have the political dynamics that can be found in the Italian third sector".
As the quote shows, some interviewees tend to differentiate the Italian context from the international context highlighting that at national level the belonging to a political culture is relevant for the career in civil society. In fact, only a few of the leaders interviewed (three) say they hold a leadership position in an organization placed "outside the culture of belonging," as stated by one of them. For these three cases the link between the previous organizations and the current organization is the sub-field and previous organizational collaborations, and this kind of changes happen more often with leaders engaged in international organizations, not always considered "part of the historical Italian third sector." This is also confirmed by the national president of an international organization active in Italy: "I'm quite embedded in the Italian third sector, I was the president of a historical Italian organization and worked in the public sector before holding this position. My current organization is international but has good relations with the Italian ones exactly for this reason, because it respects the national rules. . . I'm here for my skills and my experience, of course, but also, and I would say mainly, because I'm an Italian third sector leader, recognized at national level" In sum, the interviews highlight that in the past the third sector was strongly marked by a polarization between political cultures that can be understood as different pillars. There seems to be some consensus that nowadays these pillars are less relevant in the sector although still present. According to most of the interviewees, pillars often have a role in the career trajectory of leaders. In fact, only a small group of the leaders interviewed stated that pillars had not influenced their upward mobility in leading positions.
Our results also show that the relevance of pillars can be linked to the organizational structure. The belonging to a pillar seems to be more relevant in hierarchical and historical organizations and less in international organizations. The issue of pillars remains also controversial when it comes to relatively new umbrella organizations with members belonging to different pillars. For some leaders, the political and ideological polarization remains a specificity of the Italian third sector compared to international organizations.
In order to gain another perspective of the leaders' career trajectories and address whether the belonging to a pillar is relevant and whether its relevance depending on the organizational structure, we focus the next part of the analysis on the trajectories of 30 presidents and analyze them together with the organizational features (policy area, type, pillar, and historical development) of the CSOs they lead.

The Trajectories of the Leaders
Looking at the leaders' trajectories, we identified three different trajectory logics that link the individual leaders' trajectories with the organizational structure and identity: (1) Within logic: Multi-Level, 2) Within logic: Member organization, and (3) Outside logic: Supporting organization. Our analysis shows that these three different logics correspond to three distinct types of careers that mirror different organizational structures and are formed by the two pillars in Italian civil society. The organizations are presented in Table 1 in Appendix with information about area of activity, pillar, and trajectory logic of the president.

Within: Multi-level
The logic within: multi-level concerns 16 out of the 22 presidents who were active within the organization before holding their current position. These presidents were engaged at different levels, generally moving from the local level as activists then as representatives of the organization at the local and/or regional level to the national level and progressing from positions as national board members, directors, and vice-presidents to the position of presidents. Among these persons, we find different, specific personal trajectories but that share the multi-level experience within their organization. In this group, we can find the president of ARCI who grew up inside this organization, starting her engagement at the local level as an activist. We can also find individuals who followed this kind of trajectory and became presidents with the death of their predecessor, such as the president of ANCeScao (Associazione Nazionale Centri Sociali, Comitati Anziani e Orti) who entered this position after holding leadership positions at local, regional, and national levels for 20 years.
Looking at how much time these leaders spent in the organizations before becoming presidents, our results show that they have had long-term experience within their organizations, with most of them having more than 10 years of experience.
This logic seems to be more common for organizations with a high hierarchical and pyramidal structure. This is the case of major historical organizations that have their national headquarters at the top of the pyramid, generally based in Rome, and then have regional and local branches throughout the whole country.
As one of the clearest examples of this, ACLI (Associazioni Cristiane Lavoratori Italiani) is one of the biggest and most structurally historical Italian organizations and is active in the field of social promotion and part of the Catholic pillar. Its president's trajectory follows the multi-level logic with long-term experience as an activist. Born in 1964, he became a member of ACLI in 1994, and after a few years he became a member of the board and then president at the local level and then at the regional level. After 10 years of engagement, in 2004 he held the position of board member at the national level until becoming the national president in 2016 after more than 20 years of engagement in the organization.
Almost all of the leaders following this logic of within: multi-level have a trajectory "from below," with one exception, which is the president of the CRI (Croce Rossa Italiana). In this case, we see a similar trajectory but "from above" because the previous experiences of the president were international. Instead of moving from the local level to the national level, this president seems to have "jumped" from being engaged as a volunteer directly to gaining experiences at the international level, as international president, before becoming national president. Another interesting case that is worth highlighting is the case of the president of Legambiente, a historical Italian organization active in the area of the environment. He is an environmental engineer who also worked as an expert consultant within other CSOs and within public institutions in the field of the environment. His engagement in the organization started in 1998 as a volunteer. In 2006 he became director of the scientific department of the organization, then vice president in 2011, director in 2015, and president in 2018. For this trajectory the element of expertise, intended as technical and relevant experience in the field, seems to add on to the within: multi-level logic. However, the first logic based on a trajectory inside the organization, is also maintained in these two cases.

Within: Member Organization
The logic that we call within: member organization concerns six presidents who held positions only at the national level before being elected as the heads of their organizations. Generally, the previous positions held were positions of director, vice-president, board member, or directly of president.
This is mostly the case for the so-called networks of the second, third, and/or fourth level. As discussed, networks of the third and fourth levels are umbrella organizations that have as their members lower-level networks of organizations, while second-level networks have as their members individual membership-based organizations. An interesting case of a third-level network is AOI, the association for the Italian NGOs working with international aid. The president of this organization had 20 years of experience in different national organizations linked to or that were members of AOI.
While the organizations considered in the within: multilevel logic share the same identity as their members (as with ACLI and its regional and local chapters), in the case of these networks the representation process is more troublesome, especially when the organizations to be represented are very different when it comes to values, norms, and areas of activity. AOI, for instance, has as its members NGOs coming from both the Catholic and the communist traditions. Our short bios suggest that for these organizations, the presidents generally come from one of the most influential member organizations or have had experiences in more than one of the member organizations.
The clearest example of this is the Forum del Terzo settore. This organization is the biggest permanent civil society platform representing the whole sector at the national level and is organized as a fourth-level network that includes both second and third-level networks and first-level organizations and is thus at the top of the hierarchy of Italian meta-organizations. Its president is the vice-president of four other organizations, two of which are important members of the Forum del Terzo settore (Federsolidarietà and Confcooperative). Primarily engaged in the field of social cooperatives, this president was involved in many other influential member organizations at the local, regional, and national levels.
While this logic might appear to be based on experiences outside the organization, a closer look at the relationship between the CSO for which the leader is president and the CSOs the leader has held positions in as part of their trajectory reveals close links in terms of membership, thus suggesting a clear within: member organization logic. Thus, the leader can move directly to a position at the national level based on their experience in member organizations that are in turn national organizations. Many of these leaders have followed a within: multi-level logic in their previous experiences in the member organizations. The only exception we found in this logic is the president of Modavi, one of the biggest Italian voluntary organizations born in 1996, who has not had previous experiences in other organizations before holding positions at the national level. She is 37 years old, and is the youngest leader of our sub-sample, and before becoming president in 2014 she occupied key positions in this organization as treasurer and as a board member. She is also a board member of the National Youth Forum, a national platform of Italian youth organizations. The particularity of this case is due to the young age of the president who does not seem to have much previous experiences either inside or outside her organization, at least compared to the other presidents of our sample.
Previous research on the leadership of the Italian volunteer sector stressed that there are important differences between younger and older leaders when it comes to their trajectories before holding key positions in their organizations (Colozzi & Prandini, 2008). According to Colozzi and Prandini (2008), the majority of the leaders of the Italian voluntary organizations are more than 50 years old and have previous experiences both within civil society (other organizations, trade unions) and in other sectors such as the public sector and political parties. Unlike the older leaders, the younger leaders do not have previous experiences outside their organizations or relevant experiences inside or outside the civil society field to the point that the authors considered this aspect as a "generational variable" (Ibidem). The younger leaders might represent a "post ideological generation" (Ibidem) that accesses leadership positions without previous social and/or political engagement. In the case of the president of Modavi, her atypical trajectory-with respect to our other presidents-might also be related to the fact that Modavi is a relative new organization in the Italian civil society sphere.

Outside: Supporting Organization
When it comes to the eight presidents who did not have previous experiences within their organizations or within a member organization, we first looked at the sector where they worked before they entered the position as president of the organization. Five of them came from other CSOs, and only three of them did not have any experience in the civil society sector, two of whom came from political parties while one came from the public sector.
Based on their short bios, these presidents had no relationships with the organizations before becoming their presidents. This observation led us to ask which kind of logic had been at play concerning their link to the organizations' structure and identity, if any. In this case, the notion of pillar is particularly relevant to understanding such link. In fact, our findings show that the seven presidents had a trajectory within organizations belonging to the same pillar of the CSOs of which they held the presidency. This is, for instance, the case of MCL's (Movimento Cristiano per i lavoratori) president. This organization is clearly linked to the Catholic pillar, and its president was a member of the DC before becoming president.
In our understanding, this logic also applies when it comes to the founders of the organization (two presidents). In this case, after working in different but linked organizations for some years, the president decides to found their own organization, which will become part of their previous network. This is the case of the president of ForumSad who was the founder and president of some organizations that are now linked to ForumSad because they work in the same area of activity and pillar. In fact, this logic is based on a direct relation between the organization in which the leaders are presidents and the other organizations where the presidents were or had been engaged as part of the same organizational structure.
The president of Federconsumatori is one of the clearest examples of this logic. Federconsumatori is an organization that represents the interests of consumers and was founded in 1988 with the support of one of the most important national left-wing trade unions, CGIL, which has historically been linked to the PCI, the Italian communist party. As discussed in the theoretical section, many CSOs and civil society actors have been strongly connected to this political party and its ideological tradition. The trajectory of the president of Federconsumatori clearly coincides with that of the organization. He has been engaged for many years in the CGIL, first at the local level, then at the regional level. He was also a member of the PCI and was elected as mayor of a small town in the south of the country. Hence, this president had been involved in different social and political groups linked to the same cultural and ideological framework before assuming the presidency of Federconsumatori, an organization spawned and supported by the union adhering to the same movement. This example highlights how leaders' belonging to a specific movement or, as we call it, to a pillar (in this case the communist pillar) can play an important role in their trajectory toward the apex of power in civil society. This is not surprising given the fact four out of the seven presidents following this logic have been involved in a political party.
We also found an interesting case worth highlighting because it might be considered an exception to this third logic. This is the case of the president of Oxfam, an international organization engaged in the policy area of human rights. This president follows the Outside: supporting organization logic coming from a supporting organization, although not as clearly linked to a pillar as in the previous cases. She worked as secretary general in a national organization active in the field of international aid and human rights. With this position, she took part in international projects and committees. In 1997, she became the president of a European umbrella organization for 2 years. Both the Italian organization where she worked as secretary general and Oxfam are members of this network. After this international experience, she held the position of director in another Italian organization in the same policy area and joined an Italian leftist political party and was elected as a member of parliament. She became president of Oxfam in 2018. Her career path shows mobility linked to a specific policy field because she did not leave the areas of human rights and international development, but not as clearly as the other cases that are linked to a "pillar." Similarly to the case of the president of Legambiente discussed under the first logic, for the president of Oxfam the element of expertise also seems to have had an important role in her upward mobility to this leading position. The fact that Oxfam as an organization was established in Italy by a large international organization might partly explain why the trajectory of this leader does not fit as well with our three logics and why it is more linked to a policy area than to a pillar.
In sum, according to the type of organization, we identified three main logics: within: multi-level, within: memberorganization, and outside: supporting organization. The logics are presented in the three columns in Table 1. On the rows, we present three different dimensions in the trajectories, previous experiences within the CSO, previous experiences within members CSOs, previous experiences within the same pillar. In bold we have highlighted the main features for categorization of each trajectory in the logics. We have also assessed to what extent and how other dimensions are applicable and relevant for the leaders in each logic.
The within: multi-level logic is based on a multi-level trajectory from below and upwards within the same organization, and this is the logic followed by organizations with greater hierarchical structures and that share a similar identity. The within: member organization logic mostly characterizes the meta-organizations made up of very different organizations and is based on the president's trajectories within one or more member organizations. The outside: supporting organization logic is marked by the ideological affiliation of the president by trajectories in other organizations belonging to the same pillar as the one they are president of.

Conclusions
We set out to investigate the relation between individual leaders' career trajectories and organizational structure and identity in Italian CSOs active at national level. The interview study, focused on leaders' reflections on their own previous experiences and on the more general dynamics in Italian civil society. Although we found some differences the general picture is that the third sector in Italy was traditionally marked by cultural and ideological conflicts and hence segmented in two pillars. This segmentation affected the way in which career trajectories were possible. The interviews also suggest that the pillars have recently lost some relevance, although the relevance seem to depend on the organizational structure of the CSO. Belonging to a pillar seems to be more relevant in hierarchical and historical organizations, less so in international organizations and more controversial in relatively new more heterogeneous umbrella organizations.
In order to get a clearer understanding of the actual career trajectories of the leaders we have chosen to triangulate the interview study with the analysis of the short bios of the leaders. Here we find three different logics that mostly confirm the results of the interviews. The three main logics identified through this analysis allow us to better understand the complex interplay between individual experiences and career trajectories, organizational structures and more general ideological and cultural divides at the societal level: The first two logics are consistent with the results of our interviews. In fact, the within: multi-level logic represent well historical and hierarchical organizations in which pillars are more relevant, whereas the within: member organization logic is more linked to umbrella organizations where the belonging to a pillar is controversial. The third logic doesn't clearly appear through the interviews. However, the outside: supporting organization logic shows once again the relevance of pillars in the trajectories of leaders. Combining the results of the analysis of the interviews and of the short bios we can hence see that the issue of belonging can have two dimensions in the career trajectory of the leaders, one narrow in terms of coming from within the same organization and one broad in terms of coming from organizations affiliated to the same cultural and ideological group. All three logics tend to support our initial hypothesis of presidents' career trajectories to be at least in part shaped by the organizations' organizational structure and identity.
Through the triangulation of different data, we can see that being eligible for a position as president does not seem to be linked to a particular sphere (e.g., the field of civil society) or policy area, but rather to a specific pillar. As stated above, the history and development of Italian civil society has been strongly influenced by and connected to the two biggest political parties, PCI and DC. For many years, Italian civil society has developed around these two political parties and their ideology. As previously discussed, some recent studies have stressed that these ideological frameworks still influence Italian CSOs at the local level (Bassoli, 2013;Polizzi & Bassoli, 2011). The findings presented in this paper show accordingly that ideological affiliation can also play an important role in the individuals' career trajectory at the national level. The exceptions we found among the trajectories of the 30 presidents also confirm the results of the interview study and suggest that the organizations with an international profile might follow a logic more influenced by processes of professionalization.
Another interesting result is that time and long career trajectories seems to have played an important role in the building of the accountability of our 30 presidents. In order to be representative, the presidents have to spend a long time working within their organization or within a partner organization, preferably moving from the local to the national or international level. The Italian name for this process is "gavetta," a term taken from military jargon meaning that you need to start from the lowest ranks in the hierarchy. The clearest example is the logic within: multi-level for which the presidents become eligible for the position by the fact that they have been engaged for many years in the organization, starting from local branches before moving to the national level. These findings can be related to the relatively high age of the 30 presidents pointing to a slow reproduction process of civil society leadership. However, they are also consistent with the representations of the leaders interviewed and with the gerontocratic character of Italian elites highlighted in previous studies (Carboni, 2015). Our focus on the Italian presidents of course opens for further studies on other important leadership positions, for instance, the directors for whom we would expect a more pronounced requirement of professional, administrative, and communicative skills (see XXXX, 2020). The extent to which their recruitment also depends on trajectory logics should be further explored.
A cross-national comparison of short bios of civil society leaders would also allow for the exploration of whether different civil society regimes result in different career trajectories of the leaders. We would expect recruitment of leaders in less hierarchically structured civil society sectors to be less dependent on long trajectories within the organizations. The extent of the "pillarization" of different societies would also impact on the need for leaders' to have experiences that mirror the "right" ideological affiliations. These considerations allow us also to predict different results if we looked at career trajectories among less structured organizations in Italian civil society or at sub-national level, instead of focusing on the apex of power as we do in our paper.
A result that needs more attention in future research is the low representation of women among the leaders. The long trajectories and career paths of the leaders might suggest that these mirror the structure and reproduction of civil society leadership of past decades. The presence of women in leading positions at the sub-national level should be further explored as well as the possible presence of one or more "glass ceilings" hindering female leaders from rising to leadership positions in national peak CSOs. All in all, we argue for the need for more through studies of the processes of reproduction of the leadership looking at the way in which the different trajectory logics identified in our study might correspond to different recruitment, selection, or election processes.