Introduction

Now that governance has become a hot topic in research on civil society organizations (CSOs), it has also taken on a kaleidoscopic character. In different disciplines, theories, and cultural contexts, different notions concerning the proper addressees, contents, and mechanisms of CSO governance are used.

This study aims to explain the diverse notions of governance, by arguing that different notions of governance are rooted in different discourses of organization in civil society. To develop this argument, we examine what discourses of organization can be found in civil society (using Austria in 2008/2009 as the case in point) and what notions of governance they imply.

We begin with a literature review to highlight the main fault lines that are characteristic of recent understandings of CSO governance. Then, we explain the theoretical foundations and methods used in our empirical study. Subsequently, in order to provide a context for our empirical findings, we outline several specific features of Austrian civil society. We then present a typology of five discourses of civil society organization: managerialism, domestic, grassroots, professionalist, and civic discourse. We continue by examining what these discourses imply with regard to CSO governance. We conclude by presenting suggestions for further research.

Fault Lines in Understandings of CSO Governance

Understandings of CSO governance vary widely. In academia, unsurprisingly, the fault lines run along disciplines and theories. In CSO practice, certain notions of governance tend to prevail in certain countries, but also within countries there is large variety of notions of governance among CSOs.

At least three academic disciplines offer a particular set of perspectives on CSO governance: economics, sociology, and political science. With a few notable exceptions (e.g., Stone and Ostrower 2007), these perspectives have been used in isolation.

From an economic perspective, CSO governance is understood as a particular form of corporate governance (e.g., Jegers 2009; Speckbacher 2008). The core question is which stakeholders make valuable and specific investments into the CSO that are not sufficiently protected by contracts. It is argued that their residual rights of control should be protected by governance mechanisms such as boards, legal protection, or standardization of outputs.

The sociological perspective suggests numerous dimensions of governance structure, including formal goals, ownership, distribution of residual claims, decision-making procedures, control and accountability mechanisms, and embedded incentives (Enjolras 2009b). It is argued that governance systems of CSOs should foster collective action based on reciprocity as well as emphasizing collective ownership, democratic checks and balances, a broad range of incentives, and participatory procedures (Enjolras 2009b, LeRoux 2009).

The public policy perspective introduces the macro concept of “new” governance to emphasize the reduced influence of traditional government and the shift of responsibilities for public policy implementation to nongovernmental actors. Governance thus comprises the formal authority as well as the informal exercise of judgment by numerous actors involved in implementing public policies and programs (Lynn et al. 2000, p. 4; Blomgren Bingham et al. 2005; Heinrich and Lynn 2000; Liou 2001).

A second fault line that can be found in research on CSO governance runs between theories. In accordance with the disciplinary perspectives they stem from, different organizational theories provide different guidelines as to what governance systems should look like (Kreutzer 2009, p. 119): Agency theory (Jensen and Meckling 1976) frames governance as control in order to ensure the management’s compliance. Stewardship theory (Donaldson 1990; Donaldson and Davis 1991) starts from the opposite angle by assuming that managers do not intend to deceive stakeholders but want to do a good job and by arguing that in order to achieve improved performance, governance systems need to strengthen cooperation between boards and managers. Resource dependency theory (Pfeffer and Salancik 1978) asks which stakeholders control critical resources and suggests checks and balances between management and board. Stakeholder theory emphasizes the organization’s responsibility toward different groups within society (Freeman 1984).

When looking at CSO practice, it appears that there are considerable differences according to national contexts. If the focus of research is to be regarded as an indicator, US-American notions of governance concentrate on governance boards of CSOs and their relations with executive staff (Bradshaw 2002, 2009; Ostrower and Stone 2006, 2010; Saidel and Harlan 1998; Zimmermann and Stevens 2008). This stands in contrast to governance practice in many European countries, especially in Northern Europe, where many CSOs are democratically governed membership organizations (Enjolras 2009b, p. 769).

If one looks at the level of individual CSOs, there appears to be a broad variety of governance systems. Enjolras (2009a, b) contrasts market-based governance in board-managed CSOs with civic governance in membership organizations. LeRoux (2009) compares paternalistic to participatory governance. Contingency theory (e.g., Ostrower and Stone 2010; Bradshaw 2009) and institutional theory (e.g., Alexander and Weiner 1998) have been applied to explain the diversity of governance systems. The focus of these studies has been on board-managed CSOs, and dependent variables have mainly related to board demographics (e.g., board size, diversity, formalization, and complexity).

In this article, we seek to broaden this focus by taking the full and farraginous variety of CSO governance systems into account, including systems where governance boards do not play an important role. For this purpose, we draw on a discourse theoretical perspective, which is particularly suited for explaining more radical differences between governance systems.

Discourse Theoretical Background

We draw on a set of theories commonly referred to as discourse theory (Wetherell et al. 2001b). Discourse theoretical approaches have been used to study managerialism (e.g., Costea et al. 2008; Hodge and Coronado 2006; Hancock and Tyler 2004) and professionalism (e.g., Evetts 2003b). We extend this line of analysis by conceiving of all ways of organizing as discourses of organization.

By discourses of organization, we mean the sets of rules that constrain the forms and contents of communication about organization, by defining what is seen as meaningful in a specific community and by delineating who can communicate about what to whom (cf., Hodge and Coronado 2006). A discourse of organization thus does not designate a type of organization but a way of communicating about organization that is internally coherent and mutually distinctive. The decisive criterion is not what researchers, but what participants themselves perceive as coherent or distinct (Potter and Wetherell 1987, 170 f.).

An CSO may emulate a certain discourse of organization to a high degree (near-monodiscursivity). Typically, however, CSOs draw on various discourses and combine them (multidiscursivity, Beyes and Jäger 2005). In other words, while discourses of organization are by definition pure, real-life CSOs usually amalgamate various discourses.

The relationship between discursive and non-discursive practices is subject of major debates among discourse researchers (for an overview, see Wetherell 2001, pp. 390–392). While all strands of discourse analysis go beyond the Marxist distinction between ideology and practice (employed in classical analyses of managerial ideology such as Bendix 1956), there is considerable disagreement when it comes to details. For the purpose of this study, we find it analytically useful to use a broad notion of discourse. This is because when analyzing organizations empirically, we find that all practices of relevance are imbued with cultural meaning. We thus understand all discourse in organizations as practice, and all practices in organizations as discursive.

Discourse Analysis as Method

In empirical analysis, we follow a set of methods suggested by Potter and Wetherell (1987), Wetherell and Potter (1988), and Wetherell et al. (2001a). For sampling, we additionally draw on strategies proposed by Glaser and Strauss (1967).

In our choice of data, we needed to consider that discourse in organizations might vary depending on the context. People may talk differently in meetings than in interviews. Interviews are economical, but may be shaped by particular discourses of organization. Still, since discourses of organization are common practice in a community, it seems reasonable to assume that the discourses available to participants during interviews are the same ones that are available to them during other organizational practices. For the purpose of this article, interviews therefore appear to be sufficient.

For sampling, we started with a literature review of different ways of organizing, drawing on sources from sociology (Boltanski and Thévenot 2006; Polletta 2002; Weber 1968), organization studies (Grandori and Furnari 2008; Meyer 2005; Mintzberg 1980; Parker et al. 2007; Thornton 2004), as well as on research specifically on CSOs (Alexander and Weiner 1998; Brainard and Siplon 2004; Eikenberry 2009). On this basis, we compiled a very rough and provisional list of around 10 potential discourses of organization, thus ensuring that our research was connected to previous work (cf., Taylor 2001, p. 320).

From the beginning on, data collection and analysis proceeded in parallel. In seeking out CSOs, we relied on Internet research, databases, and personal contacts. At the beginning, we concentrated on finding CSOs that we expected to cover the complete range of discourses of organization suggested in literature. As a next step, we focused on sampling CSOs that would challenge the categories obtained from literature. After around 10 interviews, we had developed discourse categories that seemed somewhat stable. That is to say, despite our greatest efforts to conduct interviews with representatives from CSOs that would bring in further variation, ways of talking about organization began to repeat themselves. After all, in discourse analysis it is quite common to achieve theoretical saturation even after a small number of interviews (cf., Jäger and Maier 2008). We proceeded by seeking out organizations that we expected to use the discourse categories we had preliminarily established in order to check these categories against new data and further enrich our interpretations. Finally, all data were re-checked and re-coded, using the final system of categories.

All in all, we conducted 16 interviews with groups of two or more representatives of 16 CSOs, with at least one representative being a senior member of the organization. Interviews took place in 2008 and 2009. The CSOs under investigation were between seven and about 450 years old, had between zero and almost 19,000 employees, and between zero and 30,000 volunteers. Areas of activity (according to the ICNPO, United Nations 2003) covered were culture and arts, social services, sports, emergency and relief, environment, international activities, professional associations, political organizations, labor unions, religious congregations and associations, other health services, as well as other recreation and social clubs.

Interviews were semi-structured, following a funnel-shaped structure. The first question was: “If you think back to a decade ago, what has changed in your organization and what has stayed the same?” This question was followed by controlled narrative questions about the organization (who belongs to the organization, how to enter and exit, members’ rights and obligations, communication channels, positions, decision processes, written and unwritten rules, important dates and deadlines). Subsequently, specific questions were asked to test assumptions developed in previous stages of the research process. The interview ended with questions for further sampling. Interviews were scheduled to take 90 min. Depending on how talkative interviewees were and on how much time was available, interviewers could omit questions toward the end of the interview guideline in a way they saw fit for the research purpose.

All interviews were fully transcribed, using transcription conventions proposed by Wetherell and Potter (1992). These conventions are sufficiently precise for an analysis of content of discourse and broad argumentative patterns (Wetherell and Potter 1992, p. 225; Taylor 2001, p. 323). All analyses were conducted on the basis of the original German transcripts.

Interview sections quoted in this article were translated into English. We indicate such verbatim quotes by putting them under quotation marks. Abridgements and alterations due to space restrictions or anonymity concerns are indicated by square brackets. On request of interview participants, we cut out those paraverbal expressions (such as “hum,” “eh,” etc.) that do not seem relevant for the research questions at hand.

Data were analyzed jointly by both authors; interim results were frequently discussed and challenged in meetings with other academic colleagues. Discourses were identified inductively from the data, but, in order to be able to give name to what we found, we drew on established concepts (i.e., “managerialist,” “domestic,” “grassroots,” “professionalist,” and “civic” discourse). The identification of discourses was achieved, first, by isolating those text passages in which speakers distinguish between different ways of organizing, and, second, by analyzing these sections for the discourses employed. Third, after initial categories had become somewhat stable, the remaining text passages that did not involve distinctions between ways of organizing were coded to check and further enrich initial understandings. Speakers’ orientations were thus the major criterion for validating our interpretations (cf., Taylor 2001, p. 323). The process of analysis was iterative. Categories were constantly checked for coherence with new data and modified if necessary. Finally, all data were re-checked and re-coded, using the final system of categories. The final output was an analysis of discourses of organization that allows categorizing all instances of talk in our interview material and is theoretically saturated, i.e., can no longer be modified by additional observations within the basic population of Austrian CSOs.

In order to further corroborate the validity of our findings, we conducted a member check (cf., Taylor 2001, p. 322) by sending all interviewees a summary of our preliminary findings and asking them for feedback. Participant reactions to the summary were favorable.

Particularities of Austrian Civil Society

In our analysis, we use civil society in Austrian in 2008/2009 as the case in point. Methodologically, this limits the generalizability of our findings to this particular time and space. While we are confident that the discourses we identify can also be found in other settings, some aspects are bound to be attributable to the specific national context.

Austrian civil society has a number of particularities (cf., Schneider et al. 2007). One peculiarity is the importance of federalism, with central organizations in many CSOs serving only an umbrella function. Furthermore, professional organizations and interest groups are highly relevant, e.g., in the “social partnership” between employers’ and labor organizations. Moreover, the Austrian civil society has been shaped by a two-party system, with many CSOs relating either to the Social Democrats or to the Christian Democrats. In addition, many CSOs are closely connected to the Roman Catholic Church. A further particularity is the large number of small associations. The most important area of activity is that of social services, with almost 60% of all third-sector employees working there. For large parts of civil society, such as social services, arts, and culture, public funding is the main source of income.

Overall, Austria can be placed somewhere between the social democrat and the corporatist nonprofit regime. In line with the social democrat regime, Austria has a high degree of welfare, which is mainly delivered by the public sector. In accordance with the corporatist regime, civil society organizations employ a relatively large proportion (6%) of Austria’s workforce and typically co-operate with the state (Neumayr et al. 2009; Salamon and Sokolowski 2004). When it comes to volunteering, donating money, memberships, and political engagement, Austria reaches average figures if compared to other European countries (Deth 2006).

Five Discourses of Civil Society Organization

We identified five discourses of civil society organization: managerialist, domestic, grassroots, professionalist, and civic discourse. To enable a systematic comparison of discourses, we describe them along the same dimensions: important topics, modes of decision-making, constructions of actors, relationships between actors, views on what communication channels should look like, beliefs concerning appropriate personnel practices, and notions of time. These descriptive dimensions were chosen to enable a comprehensive view of the social and organizational structure suggested by different discourses. The choice was informed by social systems theory (Luhmann 1995, 75 f.; Luhmann 2003, 45 f.).

All discourses are strongly normative, i.e., they delineate how organizations should work and not necessarily how they actually do work. For example, just like managerialist ideals of instrumental rationality, grassroots ideals of egalitarianism are difficult to realize in real-life organizations.

In CSO practice, the different discourses of organization do not carry equal weight. It is probably safe to say that in many parts of civil society today, managerialist discourse is hegemonic. In the following analysis, we deliberately counteract this state of affairs by presenting managerialist discourse as one variant among others. Each discourse has its strengths, but also idiosyncrasies, which, on close inspection, could prove problematic (Table 1).

Table 1 Five discourses of nonprofit organization
Full size table

Managerialist Discourse

One discourse can be characterized as ‘managerialist’ (see, for example, Parker 2002; Pollitt 1993; Roberts et al. 2005). Its leading organizational metaphor is that of a business enterprise that produces goods and services for customers.

Central topics in managerialist discourse are effectiveness, efficiency, resources, and strategy. It is stated that the organization should choose those methods that will lead to efficient and effective mission achievement. For example, a member of a Catholic order explains: “That’s our goal: [Quotes the order’s motto.] There are different methods for doing it. […] That’s about the same as if somebody said, okay, today I write with a ball pen and tomorrow with a fountain-pen.” There is extensive talk about increasing the amount of financial, but also of human resources available, and about using those resources efficiently: “Tell the donors that they are welcome to give a bit more, because their money is well invested with us.” Speakers emphasize the need for strategy, which is considered a worthier concern than “operative,” day-to-day work.

Decision-making is supposed to follow the model of the rational management cycle: defining goals, planning on the basis of objective information and technical knowledge, implementing measures to attain goals, regularly evaluating measures with regard to efficiency and effectiveness, and making improvements.

The actors are viewed as self-interested, autonomous, instrumentally rational, and agentic. It is believed that the CSO should actively shape its relationships with others: “[…W]e want to reposition ourselves and become the leader in this issue.” It is considered appropriate that the CSO puts its self-interest of survival before certain philanthropic concerns that other actors may have. This may affect employees (“We used to have sick leaves that lasted for three years. We were a social organization. […] That’s no longer possible. The cost pressure.”) and other CSOs (“[Name of another CSO] is new to the Austrian market. […] We hope that they won’t make it. We are completely evil in this regard.”). Managerialist discourse offers a range of concepts, such as “stakeholders” or “moral owners,” to make sense of the multitude of actors that the organization is facing. Certain actors are typical of managerialist discourses: Other CSOs may be seen as “competitors,” funders are “investors,” and all sorts of actors are seen as “customers.”

Actors are perceived as relating to each other within certain markets, e.g., markets for volunteer labor, markets for charitable donations, output markets, etc. Actors expect arm’s-length exchange relationships of contributions and rewards. This creates risks and opportunities for the CSO: On the downside, with all actors pursuing their own interests, organizational unity is at threat. On the upside, the instrumental rationality of actors makes it possible for managers, who thereby assume a crucial role in the organization, to align actors’ goals with those of the organization by using incentives and installing competitive conditions on quasi markets (e.g., inner-organizational quality rankings or bonus-penalty systems). By governing motivations instead of actions, managerialism is able to tap into individuals’ agentic capacities. This is called empowerment. The freedom thus given is of a particular sort: Managers can withdraw decision-making opportunities at will, but actors are free to find another exchange relationship that suits them better.

Communication channels within the organization are carefully designed for the purpose of optimizing the tradeoff between “clarity” and flexibility. The CSO engages in activities such as “organizational development” and “restructuring.” It is attempted to formalize the reporting relationships between members, with “clarity” being a key concern. At the same time, the organization should be flexible, lean, fast, and attuned to the demands of its various market environments.

Personnel practices, including the management of volunteers, are modeled after business management ideals and focus on performance. Recruiting is supposed to follow transparent procedures, giving external and internal applicants equal chances. In personnel selection, performance potential is considered the only sensible and fair criterion. Training and development to strengthen management skills are systematically encouraged. Members are assessed against set objectives. Care is taken that members receive attractive rewards for their contributions. In case of volunteers this means ensuring the satisfaction of individual motives. In case of employees it means adequate pay, possibly including performance-based components. The CSO is aware of the benefits of volunteer work—not least of the fact that it is free labor. When it comes to deciding whether paid or volunteer work should be used for a particular purpose, the CSO has a purely instrumental attitude. If it seems efficient and effective to do so, considerable amounts of money are spent on volunteer management and development. It is considered legitimate to terminate members who do not perform, and for members to leave the organization to further their own careers.

In its notion of time, managerialist discourse is oriented toward the future. It is assumed that the organization needs to change constantly and rapidly. Members plan for the future: “We are trying to implement a three-year planning process. And on a strategic level, planning should even go beyond these three years.” There is constant and rapid organizational change, in order to keep fit for an ever-changing competitive environment. It is believed that if correct management methods are used, the future will entail progress and growth.

Domestic Discourse

In keeping with the terminology of Boltanski and Thévenot (2006), another discourse can be referred to as “domestic.” Its metaphors are those of the family and the home. Speakers characterize the CSO as a “kitchen table organization,” leaders address subordinates as their “children” or “darlings,” clients gratefully address the boss as “mama.”

Important topics of domestic discourse are everyday work and personal issues. There is much concern about “rolling up one’s sleeves and getting to work,” “keeping things running,” etc. Speakers emphasize that they prefer uncomplicated direct activity. Personal issues are central to the organization. For example, when talking about the organization, references are made to people, not to positions. Members mix CSO matters with matters from other spheres of life, e.g., by financing organizational activities out of their own pockets. People’s private problems are taken into account at work. Conflicts about factual matters quickly turn personal. “What people say” is more important than “naked numbers.”

Domestic discourse has a noticeable disregard for decisions. Instead, self-organization, spontaneity, and lucky “coincidences” are valued. People trust in God or higher laws of justice that will ensure that the organization’s good deeds are rewarded. Agency is thus largely located in exogenous forces. Orientation for decision-making is provided by a collective “spirit,” “passion,” or “idealism,” “Gut feelings” and sympathy with clients are considered legitimate decision criteria. As far as formal rules are concerned, simple rules that leave much room for “interpretation” and individual arrangements are preferred.

The organization is seen as unique, as the only one active in a particular field. Within the organization, everybody is “friends” or “family.” Actors are believed to be devoted idealists. Leaders are those persons with the most exemplary character, who do most everyday work and make the biggest sacrifices for the organization. Subordinates are grateful to leaders for shouldering the burden of these jobs. Members are committed to the organization for a long time. Clients are seen as “poor wretches” who depend on the organization’s benevolence. The organization, in turn, is believed to depend on the benevolence of benefactors and patrons. Contacts with the environment are personal and concern individuals from the local community (e.g., politicians).

When it comes to relationships between actors, “what really counts […] is the other human being.” Members are expected to make sacrifices for the organization, to “fit in,” and to behave with discretion. There are large status differences between members, which are based on the intensity and length of their engagement. Relationships between leaders and subordinates as well as between the organization and its funders are characterized not by direct exchanges, but by mutual loyalty, trust, obligation, and personal negotiations. In return for their services, subordinates can expect leaders to take care of them. A good “atmosphere” is very important. Members are “considerate of each other” and do not argue or compete. Socializing is important; people chat with each other and meet for “cozy” get-togethers. Also, local proximity is an important criterion that gives rise to friendly ties and mutual responsibilities.

Communication channels in the organization are flexible, personal, and informal. Directors and board members can be directly approached by everybody with any kind of question. Division of labor is flexible: “What needs to be done, needs to be done.” Communication often avoids official channels. Important decisions are often made in informal or even secret conversations. Formal meetings are considered a waste of time and a risk for unproductive conflicts to develop. People work to a large extent “independently”; “people’s own initiative” is highly valued. There is little separation of deliberation and action; whoever has an idea usually puts it into practice herself and works it through “on her own responsibility.”

Recruitment works via personal contacts and word-of-mouth. In personnel selection, it matters whether candidates are loyal and “fit with the group.” People who have already been known for a long time are therefore preferred. In addition, the personal circumstances of loyal members may be taken into account. High qualifications and performance are regarded as less important or even disruptive to organizational harmony. If elections take place within the organization, in contrast to those in civic discourse, they are not political affairs. Instead, candidates are asked whether they would like to take over a position and “nobody really says no.” Elections can then be unanimous or even by acclamation. Learning takes place informally, through learning by doing, watching others, or by having “people explain things to you, and then you read for yourself.” In assessment, everybody is measured against their individual possibilities. Monetary rewards are perceived as somewhat mundane or even immoral. In some sense, members are expected to work for no reward, out of a sense of duty or “social engagement.” The only worthy rewards are the gratitude of clients as well as “fun” and conviviality at work. Accordingly, superior worth is attributed to volunteers. People usually only leave the organization when they retire or die, or because of personal conflicts with other members.

In its notion of time, domestic discourse emphasizes the past and present. The organization’s past is well remembered; traditions are preserved. Present and short-term needs are in the focus of attention. Spontaneity and quick reactions are important; plans are always open to revision. The pace of activities is erratic: While usually the pace is leisurely, sometimes things become chaotic, with a sense of teetering “on the brink of disaster.”

Professionalist Discourse

A further discourse may be labeled “professionalist” in the sense of substantive professionalism such as traditionally fostered by the medical and legal professions (cf., Freidson 2001; Evetts 2003a). The organization is depicted as a pool of experts who use their discretionary knowledge to solve complex problems.

The main topics are the challenges and quality of the organization’s substantive work. For example, when asked to outline how the organization had changed in the previous decade, an interviewee from an environmental organization gave a 5 min speech on the changed nature of ecological threats. Similarly, the head of a student exchange organization gave us an account of the changed geopolitical situation and its implications on visa issuance.

Decision-making is decentralized because work is regarded as discretionary and complex. Staff members receive little immediate supervision or feedback from supervisors. Instead, their work is guided by ideals and standards that originate from their profession. Staff members are not purely results-oriented but committed to ethics of the field (e.g., “fair play” in soccer). Services are offered not because of market demands but because of “a certain substantive conviction and stance.” The quality of work is paramount, which entails a strong concern for safety. Speakers approve of using available resources efficiently, but there is little awareness of the possibility to increase resources by entrepreneurial means. In some cases, the commitment to quality may lead to conflicts with cost efficiency.

Actors are defined by their profession. A key distinction is the one between experts and laypersons. Professional identity is strong; the members of a profession have a shared understanding of their work that is grounded in shared knowledge and a common educational background. Organizational identity, in contrast, is often weak.

When it comes to relationships between actors, the focus is on relationships among experts and relationships between experts and clients. Among experts, there is a great deal of co-operation with colleagues from the same specialty who work outside the organization. This is because all are viewed as working for the same higher purpose. Other organizations in the field are regarded as colleagues, partners, and “friends.” Within the organization, there may be conflicts between different professions because of divergent perspectives. The relationship between experts and their clients is “professional,” i.e., characterized by distance rather than empathy. This means that staff members have a rational approach to work, even if it is emotionally challenging. Acknowledgment by peers and succeeding in competitions are considered important criteria for success. Status differences between staff members are based on differences in knowledge and qualification. All members take great pride in “accomplishments” (“We want to proof that we can do it.”).

Communication channels extend beyond the organization into the individual’s profession. There is much teamwork between staff members of the same profession. For the sake of quality, there is a fair amount of record keeping and reporting. Positions or departments that oversee the substantive aspects of work are powerful within the organization. Departments that handle organizational aspects play an unimportant role and carry unassuming names, such as “accounting,” “administration” or “commercial directorate,” and do not interfere with the work of substantive professionals (“That requires special knowledge.”). Substantive professionals, in contrast, are involved in all important decisions (“We do pedagogic work. […] And therefore pedagogics is so important and involved in everything”). “Administrative” or “commercial” positions are regarded as a practical necessity and often rotate, i.e., substantive professionals take over these positions for a limited period of time.

With regard to personnel practices, professionalist discourse emphasizes educational achievements and “proficiency.” Staff members are preferably recruited from outside the organization via transparent procedures. Proficiency and educational achievements are the central selection criteria. Staff members are well-trained (e.g., hold a relevant university degrees or have received “the fire fighters golden proficiency badge”). Formal education outside the organization plays an important role and is valued per se, not just as an economically rational investment. People are intrinsically motivated to specialize and follow latest developments in the field. The quality control mechanism for daily work is self-assessment against professional norms. At large intervals, performance is assessed through formal examinations and peer reviews. These usually take place in the professional community beyond the organization. When it comes to rewards, it is believed that staff members should receive an adequate, fixed salary. There is a positive attitude toward paid work; it is considered only fair that qualified work should be remunerated. Being recognized as an expert by peers and laypersons is also considered an important reward. It is believed that staff members should have a reasonable work-life balance; self-sacrifice is neither expected nor valued. Volunteering is conceived of as work and not “just a hobby.” Even volunteers strive to be “as excellent […] as a professional team.” Legitimate exit from the organization can occur when moving on to another organization in the same professional field with the aim to further develop one’s skills.

The notion of time is future-oriented, with a controlled pace of work. It is believed that the field is constantly changing, which requires the organization to stay up-to-date. Improvement and learning are therefore encouraged. Staff members take pride in being “cutting edge.” It is, however, believed that improvements need time; there are no quick and easy solutions. The desire for quality leads to a preference for organizational stability or slow sustainable growth.

Grassroots Discourse

Another discourse can be characterized as “grassroots” (see, for example, Brainard and Brinkerhoff 2004; Smith 2000). The central notion of grassroots discourse is that the organization should be a domination-free space.

The main topics are the organization’s “principles” and “positions.” “Principles” are keywords that are known to all members (e.g., “subsidiarity” and “nonviolence”). “Positions” are more elaborate and define the organization’s stance toward substantial matters. For example, an antifascist organization may have a position on: “When is a statement a relativization of the horrors of the holocaust?” Since the organization’s goals are typically abstract and difficult to realize, being true to one’s principles is considered an indicator of success.

Decisions are made by consensus. If no consensus is found, one way out is to find a consensus that regulates how to further deal with the problem, for instance, by carrying out a majority vote. This means that every individual has a veto right on every decision. Due to their shared idealism, members exercise this right “responsibly.”

Actors in grassroots discourse are constructed as autonomous. For members, autonomy means that they are fully informed about all issues, participate in decisions, know why a particular decision has been made, and, consequently, fully support the organization’s course of action. This autonomy is not just a right but also a responsibility. Every member is personally responsible for all decisions. Individual members have a responsibility to argue their point. This puts high demands on members: They have to deal with different opinions, argue their way through controversies, and come to joint decisions. The organization as a whole takes care to remain autonomous from funders. This can be achieved by keeping financial needs to a minimum, i.e., by using volunteer work only. Alternatively, the organization may accept grants that come with no strings attached.

Relationships between actors are characterized by egalitarianism and collectivism. Egalitarianism implies that hierarchies are rejected; everybody has an equal say on everything. Even hierarchies based on different levels of knowledge are viewed with suspicion. Ideally, everybody should be competent in everything. There are no official leadership positions; implications of hierarchy are avoided in job titles. Often all members are simply referred to as “people,” “persons,” “women” (in case of feminist organizations), or “activists.” The organization has elaborate rules and tools to ensure equal participation (e.g., quotas, “lists of speakers”). There is high sensitivity for gender issues. The organization deals with its environment in ways that reflect collectivism. It chooses groups instead of individuals as representatives. These representatives rarely have a “negotiation mandate” but merely collect information and report back to the plenary. The plenary decides, anticipates possible developments, and sends the representatives back to continue negotiations.

Communication channels in the organization aim for maximum participation and openness. It is believed that as many people as possible should participate in every decision and activity. The organization is typically steered by a “collective” or a “plenary,” where all members jointly decide on all organizational matters, including operative work. This means that the organization must not be bigger than its plenary, which implies small organizational size. Interviewees emphasize that their organization aims to be open. It is easy to be admitted into the organization as a member. Sometimes this leads to ill-defined organizational boundaries. Within the organization, there is wide sharing of knowledge, especially by means of oral communication. Written records are accepted only insofar as they remain flexible and open to renegotiation (e.g., Wikis, minutes). Records are accessible to all members. Elaborate reports to the public, in contrast, are regarded as self-aggrandizement, manipulation, and bad use of time. However, it is claimed that on inquiry, the organization is totally open and honest to outsiders.

Members are recruited by means of open invitation to participate. Ideally, people volunteer their services on their own initiative. There is no formal selection procedure; people self-select on the basis of their identification with the organization’s principles. Members develop their skills by means of simple methods, such as learning by doing, learning from organizational records, or asking around. There are low requirements when it comes to members’ performance in a managerialist sense. A certain level of activity is all that is required. Members’ personal quirks or even mental problems are to a large extent tolerated. As grassroots discourse is highly idealist, members are expected to devote large amounts of time and to work for the greater good instead of their private interests. There is little to gain from working for the organization, except for the satisfaction involved in having done something valuable and in having had a positive growth experience. Volunteer work is the ideal. Members may, however, receive a financial allowance or stipend, not in the sense of pay but to enable them to continue working for the organization. Financial allowances are egalitarian or needs-based. It is considered legitimate for members to leave the organization if their life circumstances no longer permit engagement. Members may also be expelled if they violate organizational principles.

The notion of time focuses on the present. The organization’s strength is spontaneous, short-term activism. At the same time, organizational activities are dominated by long discussions, which are believed to be the solution to any kind of organizational problem. Certain issues are discussed over and over again to bring new members on par and also include their views.

Civic Discourse

In line with Boltanski’s and Thévenot’s (2006) concept of civicness, a final discourse may be referred to as “civic.” Here, the organization is constructed as a res publica. Positions, units, and practices within the organization have similar names like those in governments and administrations (e.g., “officer,” “commission,” and “resolution”).

Important topics are those of mass support and proper procedures. Speakers emphasize their organization’s “broad” membership base and support within the population. Much time is dedicated to talking about proper, formal, written procedures.

As far as decision-making is concerned, elaborate written rules, elections, and consensus are crucial. Written rules are believed to guarantee “clarity” and fairness. The law, bylaws, and organizational policies are taken seriously and are considered as useful (“These are the official regulations of the fire brigade. […] That’s sacred to me, nobody is allowed to touch it.”). Membership rules, rights, and responsibilities, as well as communication channels, hierarchies, and other policies are clearly defined. Meetings are carefully documented in minutes. From the point of view of other discourses, the many regulations of civic discourse seem “dry,” “unnatural,” and bureaucratic. Elections are the typical method employed to fill both leadership and supervisory board positions. The power base of leaders lies in the popular support that they have. Depending on whether they enjoy unitary or only partial support, they can then decide more or less autocratically within the scope of their office. Consensus, which assembles the support of the largest possible number of people, is the preferred mode of decision-making. However, while in grassroots discourse the purpose of consensus is to guarantee the autonomy of individuals, here the purpose is to unify and strengthen the organization in the face of external actors. If consensus is not attainable, majority decisions are accepted.

Actors are expected to display good citizenship. It is assumed that the CSO should be a membership organization, typically an association, in which all members are fundamentally equal and have active and passive voting rights. Hierarchy and differentiated participation rights are, however, accepted insofar as they are based on universal rules and democratic procedures. Executives are either elected officials or appointees. All members of the organization are expected to uphold civic virtues, notably to act in the interest of the greater community. Speakers take pride in the fact that their organization provides public goods for free.

When it comes to relationships between actors, the organization is highly conscious of diversity issues, differences of interest, and power struggles. Civic discourse accepts and even welcomes diversity within the organization, which corresponds to the wish to secure a broad membership base. It is believed that the organization should reflect and represent all relevant groups (e.g., federal states and political affiliations). Differences of interest within the organization are seen as natural. Conflicts are accepted and worked out actively and openly by means of formal procedures. Such procedures include long meetings with “heated debate” and possibly majority votes, arbitration boards, and “disciplinary proceedings.” Interviewees frequently frame relationships between actors within the organization, or between the organization and its environment, as power struggles.

Communication channels are tailored to maximize democratic legitimacy as understood in representative democracy. The organization comprises a sophisticated system of checks and balances, with many layers and a clear division of responsibilities. Much thought is spent on the composition of decision-making bodies and the majorities or minorities that can be found there. The basic governance structure is circular, i.e., members elect the leaders, and these leaders are then allowed, within the scope of their office, to give orders to the basis. Decision-making is basically bottom-up, with the organization structure ensuring that members and local chapters are the most influential players within the organization. Accounting, budgeting, and reporting to the membership base and public are considered essential for transparency purposes, not for maximizing efficiency or “presenting oneself.”

Personnel practices are concerned with issues of fairness, clarity, and representativeness. Transparent recruitment procedures are important, no matter whether candidates are recruited from within or from outside the organization. Leadership positions are considered as desirable positions that many members would like to hold. Consequently, the distribution of these positions sometimes involves power struggles that involve rival candidates, crucial votes, and voting people out of office. For selecting personnel, qualifications and representativeness of important groups are equally important criteria. In some cases, the former criterion may conflict with the latter. Civic discourse values internal training. The central methods for assessing officials are (re-)elections. Appointees may be appointed and assessed by a jury. Voluntarism is highly valued per se, not just as an instrument for other purposes (“[…W]e are a volunteer organization, it’s a volunteer culture. […W]e are proud of it and promote it very deliberately. […]”). Volunteers may receive reimbursements of costs. Paid staff members typically receive fixed pay. Employment with the organization is usually for the long term; the legitimate exit for appointed staff members is retirement. Elected officials may be voted out of office.

In civic discourse, the notion of time emphasizes stability. It is believed that the organization’s structure with its system of checks and balances is so sophisticated that it will endure time. Organizational activities, such as meetings, are performed at a slow pace, take place at fixed times, and follow a fixed structure. The organization engages in long-term planning under stable expectations concerning its environment. It engages in little of what managerialist discourse would call “agenda setting.”

Implications for CSO Governance

Each discourse of organization has distinctive governance implications. We describe these implications by examining each discourse’s answers to three core questions of governance: “To whom are we accountable?”, “For what are we accountable?”, and “How can we ensure accountability?” These questions derive from the understanding that any governance system consists of structures and processes to ensure the organization’s performance accountability to relevant actors (cf., Stone and Ostrower 2007) (Table 2).

Table 2 Governance implications of different discourses of nonprofit organization
Full size table

Accountability to Whom?

All governance systems have to answer the question: “To whom are we accountable?” (cf., Stone and Ostrower 2007, p. 423). Stakeholder theory proposes that a CSO is accountable to those actors who have contributed specific and valuable resources but whose claims are not sufficiently protected by contracts (Speckbacher 2008, p. 302). If we look at this proposition from a discourse theoretical perspective, it is apparent that the construction of actors, their contributions, and their residual claims lie within the organization. Depending on the specific type of discourse, there will be different views as to which actors the governance system should protect most.

In managerialist discourse, donors and funding institutions are central: “I always want to be able to tell the donor, with pride, that we have invested your money well.” This is because funders’ contributions are viewed as highly similar to owners’ equity in business corporations.

In domestic discourse, the organization’s prior accountability is to beneficiaries (“the kids,” “the families,” “the people who need our help”). This is because beneficiaries lack representation at the family table, at which employees are core participants.

In professionalist discourse, the organization’s prior accountability is to those who represent professional standards. For example, a fire brigade may be considered accountable to the provincial association of fire brigades as far as abiding by technical guidelines is concerned; a social services CSO may be considered accountable to the inspecting authority.

Grassroots discourse avoids external dependencies and prioritizes accountability to activists, who contribute their work to the organization. For example, in an antifascist CSO, “if a [local] group formed that […] wanted to espouse the principle of fascism, that would have to be permitted. The idea is that they have to be able to do as they please.”

Civic discourse implies a membership-based CSO and stresses democratic rights, which is why it emphasizes accountability to active members: “The president with his budget is accountable toward the club, of course. That is to say, at each general assembly of members, the president and his treasurer have to disclose the budget, discuss it, and hold a vote on whether they should be discharged from their liabilities.”

Accountability for What?

A second crucial question for governance is: “For what are we accountable?” (cf., Stone and Ostrower 2007, p. 423). Generally speaking, the CSO is accountable for its performance. However, depending on the specific type of discourse, actors have markedly different views about what performance means.

In managerialist discourse, performance means the effective and efficient achievement of an explicit mission: “Using our main operationalized goal as an indicator, I may say that we have been very successful. […] With regards to climate change, an important substantial indicator has been the passing of a climate change law. And of course we have also defined indicators in other areas that depict organizational goals.”

In domestic discourse, performance also means achieving a mission, but in contrast to the way in which the term is used in managerialism, here the mission is understood as an ideal that is shared intuitively: “We cannot measure our success, because we just try to somehow help the people. […] Success simply means success in humanitarian terms, if we can just help the people a bit with their life, in the short or medium run.”

From a professionalist perspective, performance means meeting professional standards and receiving successful peer evaluations. For example, for firefighters, achieving a low “time to scene” is considered a criterion of success: “We have achieved a standard that can no longer be surpassed. At the command center, we are known for being the fire brigade that moves out fastest.”

In grassroots discourse, the CSO is seen as successful if it stays true to the principles of grassroots democracy: “It is in the nature of things that progress is slow. […] To a large extent, [the organization] often simply tries to exemplify things through our own activities. For example to show that grassroots democratic structures are possible by structuring ourselves that way […].”

In civic discourse, it is believed that the most valid criterion for performance is the ability to secure mass support: “Election results are […] directly measurable success.” “The hard facts are the number of members, the amount of membership fees collected, the number of works councils, and the number of organized businesses.”

How to Ensure Accountability

The third crucial question for governance systems is: “How can we ensure accountability?” This question addresses the structures and processes that characterize the governance system.

Managerialist discourse promotes a business-like governance system, with boards and executive directors as the main actors: “[… As a board] you write policies and only take on strategic responsibilities and no longer do operative work. And you do monitoring. […T]he board talks to the [executive director], the [executive director] talks to the rest.”

Domestic discourse relies on personal relationships (“people are really grateful”) and inner feelings to ensure accountability: “[…F]or a development aid worker it is quite something if he is allowed to stand there […] and suddenly he is obliged to work with the money in alignment with his soul, and then he realizes how wonderful that actually is.”

In professionalist discourse, accountability is achieved within the profession, via peer assessment and comparison with other organizations in the field: “Every November we have a concert assessment […]. This is a provincial scoring system, run by the Brass Music Association and the province of Lower Austria. They fund according to the achieved amount of points. […] And of course it is also interesting if […] they say: ‘That was not okay, because of this and that.’”

In grassroots discourse, activists’ accountability toward each other is ensured by rules of grassroots democracy, notably domination-free discussions and consensus-seeking. For example, if people in a feminist organization are faced with an activist who does not use gender-sensitive language, “the approach is to explain to people why we see it as a problem.” Through discussion, a “learning process” is initiated in the person, or the organization reaches a new consensus about the issue. The danger of indulging in too much organizational self-contemplation is counteracted, grassroots discourse claims, by organizational openness: Anybody concerned about how things are going in the organization is free to join and contribute their critical view.

In civic discourse, accountability toward the membership base is assured by following rules of representative democracy, such as elections and votes, checks and balances, and adherence to formal rules: “We have to conform to accountability requirements under commercial law. […] Then there is the statute, the election regulations, and the bylaws. We have guidelines for administration. We have a signature regulation. We have a budgeting handbook. We have descriptions of work processes. […] We have travel regulations. […] And petitions, resolutions and minutes […] that have been decided on politically to determine the direction for the next years.”

Conclusions and Suggestions for Further Research

We have sought to extend the understanding of CSO governance by discussing a larger and more variegated range of CSO governance systems. Our findings show that different discourses of civil society organization give specific answers to questions of governance. Thus, our study contributes to what, in the literature review, we have labeled a sociological perspective on CSO governance. In line with this perspective, we contend that CSO governance occurs in numerous dimensions of organizational structure, and that there are variants of governance that differ from managerialist discourse and may be considered distinctive of CSOs.

Against the backdrop of our findings, it appears that in many cases, academic research about CSO governance remains within the confines of particular discourses of organization. Most prominently, research from the economic perspective is rooted in managerialist discourse, usually without displaying much reflexive awareness of this fact and of alternative discourses. Research from the sociological or political science perspectives, in contrast, often positions itself within civic discourse and sets itself apart from managerialist discourse (LeRoux 2009 and Enjolras 2009a, b can be read as examples of such a positioning.). While it is clear that academic research is not independent from the boundaries of societal discourses of organization, we think that research on CSO governance would benefit from becoming more reflexive about the discursive boundaries within which it operatives (thus taking up a quality criterion of discourse analysis, cf., Taylor 2001).

Our analysis leaves crucial issues of power and historical context open to further inquiry. Managerialist discourse has not always been as hegemonic as it is today, and alternative discourses have probably seen better times. In order to put today’s understanding of apt ways of civil society organizing and governance into context and to open up new perspectives, further research on the development of discourses over longer periods of time would be valuable. With regards to the issue of power, two important aspects warrant further inquiry: First, more research is needed as far as the power of different discourses of organization is concerned, in order to investigate how discourses influence individual and collective consciousness, subjects, and action. Second, further research should tackle the question of power over discourses of organization and examine the various ways in which different individuals and groups have different chances to influence discourse.

A further interesting objective for future research would be to examine why, how, and with what results real-life CSOs mix and match elements from different governance systems. From our empirical analysis, we have gathered the impression that each notion of governance can become dysfunctional when taken to extremes, and that the more successful CSOs are those that combine and balance the rationalities of several discourses of organization (a notion similar to the one expressed, for example, by Grandori and Furnari 2008). More empirical research is needed to investigate this observation.

Further research is also needed to examine and broaden our findings. First, since the validity of our empirical findings is restricted to civil society in Austria, further research is needed to develop typologies of discourses of organization that apply to other national contexts. It would be instructive to conduct studies of CSOs in several countries and develop a typology of discourses of civil society organization that is internationally valid. Also, in order to get a full-fledged view of governance practices in CSOs, further research that draws on data beyond interviews would be desirable.