Islands of Deliberative Capacity in an Ocean of Authoritarian Control? The Deliberative Potential of Self-Organised Teams in Firms

Abstract

Business firms play an increasingly influential role in contemporary societies, which has led many scholars to return to the question of the democratisation of corporate governance. However, the possibility of democratic deliberation within firms has received only marginal attention in the current debate. This article fills this gap in the literature by making a normative case for democratic deliberation at the workplace and empirically assessing the deliberative capacity of self-organised teams within business firms. It is based on sixteen in-depth interviews in six German firms which practice various forms of self-organised teamwork. The article argues that self-organised teamwork can create a space for authentic, inclusive, and consequential deliberation by suspending authoritarian control structures within business firms. Finally, the article proposes the consideration of firms not only as necessary parts of a larger deliberative system but also as deliberative systems in themselves.

In recent years, the question of the democratisation of corporate governance has returned. There have been proposals to democratise business firms from various perspectives, including socialism, liberal egalitarianism, neo-republicanism, or pragmatism. ¹ Similarly, there has been a great variety of arguments for workplace democracy, including relational equality, meaningful work, non-domination, democratic education, and the state-firm analogy (Frega, Herzog, and Neuhäuser 2019). Yet, the possibility of democratic deliberation within firms has gained only marginal attention among political theorists and business ethicists (Felicetti 2018). So far, the most common approach to the deliberative democratisation of corporate governance has focussed on stakeholder deliberation (e.g., Fung 2003; Palazzo and Scherer 2006; Goodman and Arenas 2015).

This article fills this gap in the literature and takes up Felicetti’s (2018, 812) call for “a tighter connection between empirical studies and theorizing on democracy in firms.” I argue that self-organised teams within firms exhibit a high level of deliberative capacity and can create spaces for authentic, inclusive, and consequential deliberation (see Dryzek 2009). I underpin these claims by in-depth interviews in firms that practise various forms of self-organised teamwork. This methodological approach is inspired by empirical studies of deliberation with interpretive methods (e.g., Parkinson 2006; Hendriks 2011; Felicetti 2016) and studies in business ethics that use qualitative methods to study ethical questions in firms (e.g., Toffler 1986; Jackall [1988] 2010). On a theoretical level, I adopt Felicetti’s (2018, 810) view that firms should be considered as necessary parts of a larger deliberative system. Furthermore, I follow the view that we should think of firms as deliberative systems in themselves (Sabadoz and Singer 2017; Singer 2019; see similarly Tamura 2014). This perspective allows us to understand how smaller organisational units of firms (e.g., self-organised teams) can exhibit a high deliberative capacity even while other parts remain non-deliberative.

The article is structured as follows. In the first section, I briefly retrace the development which led to the “systemic turn” in deliberative scholarship. On the basis of this outline, I discuss four reasons why firms should be considered as necessary parts of a larger deliberative system and why it is helpful to think of firms as deliberative systems in themselves. This is followed by an overview of the methodological approach. The next section provides a summary of the origins of self-organised teamwork and introduces the concept of deliberative capacity. Then, the results of the study and their implications for a deliberative democratisation of firms are presented. I conclude by sketching a proposal for a three-level approach to workplace deliberation.

THE DEVELOPMENT OF THE DELIBERATIVE SYSTEMS APPROACH

Today, deliberative democracy is one of the most important fields in democratic theory, with vast numbers of theoretical and empirical studies (Dryzek and Niemeyer 2010; Bächtiger et al. 2018). One can distinguish three stages in the development of contemporary (Anglophone) deliberative democratic theory.

Stage 1: The Theoretical and Normative Foundations of Deliberative Democracy

The first stage was shaped by the development of the theoretical and normative foundations of deliberative democratic theory (Elstub 2010). A starting point was Mansbridge’s (1980) distinction between “adversary” and “unitary” democracy, that is, democratic decision-making based on preference aggregation or preference transformation and consensus, which marked the beginning of the so-called deliberative turn in democratic theory (Bohman 1998; Dryzek 2000). Around the same time, legal scholars (e.g., Joseph M. Bessette, Cass Sunstein, and Bruce Ackerman) began to develop models of deliberative democracy (Bohman and Rehg 1999; Floridia 2018). ² Following their approach, scholars began to reflect on the ideal conditions for democratic deliberation and fine-tuned the deliberative conception of democratic legitimacy (e.g., Manin 1987; Cohen [1989] 2003).

An important influence on the first stage of deliberative democratic theory was Jürgen Habermas ([1961] 1989), especially his conception of the public sphere, his concept of communicative action, and his notion of the “ideal speech situation” (Habermas [1981] 1984). In such a situation, deliberation ideally achieves an un-coerced consensus by the “force of the better argument” (Habermas 1996, 306). Another important influence was John Rawls (esp. Rawls 1993), whose concepts of public reason and public justification were taken up by deliberative democrats (Chambers 2018).

By the late 1980s, deliberative democracy had developed into a mature theoretical approach with a variety of competing conceptions of deliberation. Common to them is the idea that democratic legitimacy derives from the free and public deliberation of equal citizens. Also, it is possible to identify certain “standards for good deliberation,” including respect, absence of power, equality, the use of reason, the aim at consensus, common-good orientation, publicity, accountability, and sincerity (Bächtiger et al. 2018).

Stage 2: The Expansion and Critique of the Deliberative Ideal

The second stage can be characterised by the expansion and critique of these deliberative standards, along three dimensions. First was an expansion of what should count as deliberation in pluralistic societies (Elstub, Ercan, and Fabrino Mendonça 2016). Scholars, such as Fraser (1990), Benhabib (1992), Young (2000), and Deveaux (2003), began to problematise the overly rationalistic standard for deliberation, arguing that it should be open to other forms of communication to guarantee the inclusion and political equality of all individuals and groups (Young 2000, 53; see similarly Sanders 1997, 371). In addition, the orientation of deliberation to the common good has been criticised (Fraser 1990). This critique has led many deliberative democrats to adopt more flexible, inclusive, and pluralistic conceptions of deliberation and consensus (Chambers 2003).

Second was an expansion of the idea of the public forum. Again, Fraser (1990, 76–77) contributed to this expansion by arguing for a “multiplicity of publics,” which would allow for the “autonomous opinion formation [to be] removed from the authoritative decision-making” (see similarly Dryzek 1990, 2000). Third was a move towards empirically orientated research (see Dryzek and Niemeyer 2010; Owen and Smith 2015). Scholars like Dryzek (1987) and Fishkin (1991) turned to questions of institutional design, implementation, and experimentation, putting a stronger emphasis on testing the “feasibility” of the normative principles of deliberative democracy (Elstub, Ercan, and Fabrino Mendonça 2016, 4). This “empirical turn” drew criticism from political science and psychology (e.g., Mutz 2008; Sunstein 2002; Bagg 2015; for a review of the debate, see Rosenberg 2014) but also found defenders who presented evidence of high-quality deliberation (unbiased, open-minded, reasonable, unpolarised) (e.g., Sulkin and Simon 2001; Fishkin and Luskin 2005; Strandberg, Himmelroos, and Grönlund 2017; for a discussion, see Ryfe 2005). Thus much depends on how the group context is designed and if it encourages adoption of a deliberative attitude (see Ryfe 2005; Mercier and Landemore 2012). A second type of critique focussed on the “institutional turn,” arguing that “small-scale deliberative experiments” received too much attention in comparison with attempts to make “mass democracy itself more deliberative” (Chambers 2009, 331). The question of how to integrate mini-publics in the macro-political landscape sparked controversial debates (Goodin and Dryzek 2006; Lafont 2015), which contributed to the “systemic turn” (Elstub, Ercan, and Fabrino Mendonça 2016).

Stage 3: The Deliberative Systems Approach

While one can trace similar ideas to earlier scholars (e.g., Fraser 1990; Gutmann and Thompson 1996; Dryzek 2000), the first explicit use of the term deliberative system can be found in Mansbridge (1999), where the author develops the idea that a comprehensive deliberative system encompasses not only deliberations in public decision-making or the broader public sphere but also citizens’ everyday talk. In her model, the different parts “mutually influence” each other and contribute to different degrees to the deliberative quality of the system as a whole (Mansbridge 1999, 213). A fundamental aspect of the deliberative systems approach is the division of labour (Elstub, Ercan, and Fabrino Mendonça 2016, 6). Accordingly, not all parts of the system need to be “deliberative” (Mansbridge 1999, 224). ³ The idea of a division of labour among the different parts of a system diverges from the view that only specific kinds of institutions or spheres can (or should) offer the necessary conditions for authentic deliberation (Habermas [1981] 1987, 2:152). In contrast, the deliberative systems approach assumes that deliberation can be realised in distributed ways in various spheres and institutions (Mansbridge et al. 2012, 2; see also Chambers 2017).

Another characteristic feature of the deliberative systems approach is a contextual understanding of deliberation. For example, in a highly polarised setting, even a small “signal of willingness to talk” can be considered as contributions to the deliberative quality (Parkinson 2018, 434). Another contextual factor can be “internal constraints,” which are necessary for an institution to fulfil its function (Mansbridge et al. 2012, 12–13). This might apply to firms in the sense that they need to be economically efficient.

Although there is no unitary measure for deliberation, Mansbridge et al. (2012, 11) propose the evaluation of a deliberative system or its parts according to the extent to which three functions are fulfilled. First is an epistemic function, understood as the ability to generate opinions and decisions that are informed by facts and meaningful reasons. Second is an ethical function, understood as the ability to promote relationships of mutual respect among citizens and recognising them as self-authoring sources of reasons and claims. Third is a democratic function, understood as the inclusion of plural interests and concerns on the basis of equality.

A SYSTEMIC PERSPECTIVE ON DELIBERATION IN FIRMS

From a systemic standpoint, it may seem unnecessary to make firms more deliberative, as not all parts of a deliberative system need to be deliberative (Felicetti 2018, 811). However, I will follow Felicetti (2018, 811), who argues that the “deliberative reform of firms” is a crucial factor for “the overall ability of a system to be deliberative.” There are at least four reasons why firms should be considered important parts of a larger deliberative system.

Firstly, most adults spend half their waking time at their workplaces, which makes it “the single most important site of cooperative interaction and sociability among adult citizens outside the family” (Estlund 2003, 7). Thus many of the conversations—and potentially deliberations—in which citizens engage on a daily basis take place in the workplace. Furthermore, at the workplace, employees “must find ways of cooperating on an ongoing basis,… outside of and often counter to traditional racial, ethnic, or sexual hierarchies” (Estlund 2003, 12; similarly Perry 2014). As Felicetti (2018, 809) points out, firms represent a unique environment for democratic deliberation which could contribute to the social cohesion of a democratic society and should be seen as an important part of the deliberative system.

Secondly, the structures of many workplaces are in dire need of ethical reform. Today, most firms are governed by profoundly inegalitarian or undemocratic norms (e.g., Landemore and Ferreras 2016; Anderson 2017). Employees often have little or no say in decisions that concern the firm’s strategy or its internal organisation. Furthermore, the authority of employers is accompanied by pervasive control that can extend even beyond working hours (Anderson 2017, 38–39). Such unaccountable power can lead to abuse, with significant “costs to workers’ freedom and dignity” (Anderson 2017, 71; from a republican perspective, see Hsieh 2005; González‐Ricoy 2014; Breen 2017).

Thus the organisation of most workplaces is at odds with the ethical and democratic functions of a deliberative system (Felicetti 2018). Most firms neither promote relationships of mutual respect nor guarantee the inclusion of plural interests and concerns on the basis of equality (Mansbridge et al. 2012). Thus, if deliberative democrats have as their goal the broad implementation of the ethical and democratic functions, they should aim at making firms an integral part of the deliberative system.

Thirdly, firms and their decisions play an influential role in contemporary societies (Scherer, Palazzo, and Matten 2009; Chandler and Mazlish 2005), with firms exerting significant influence on political decision-making (see Barley 2007; Lessig 2011). Economic globalisation brought an erosion of the traditional “division of labor between the political and economic spheres” (Scherer, Palazzo, and Matten 2009, 327). In today’s globalised economy, transnational corporations participate in political processes and are able even to define economic rules, regulations, and standards (Fung 2003; Young 2004). Furthermore, they are increasingly involved in providing public goods and services (Kaul et al. 2003).

In light of these developments, firms can no longer be considered purely private entities but should instead be perceived as organisations which “transgress the public/private divide” (Ciepley 2013, 152; see also Chandler and Mazlish 2005; Néron 2010). Yet the growing political role of firms has been accompanied by a democratic deficit as corporate decisions are neither democratically legitimised nor controlled (Orts 1995; Kobrin 2009; Scherer, Palazzo, and Baumann 2006). This democratic deficit provides a third reason to make firms more deliberative.

Fourthly, I agree with Felicetti (2018, 806), who argues that deliberation within firms could provide valuable spaces for critically reflecting on the current production conditions and their influence on the distribution of income and wealth in contemporary societies. Such reflection is particularly important if we assume that social inequalities can undermine the equal access of citizens to participation in political practices (Young 2000).

These four reasons should lead deliberative democrats to consider firms as important sites for deliberation in a larger deliberative system. Deliberation in firms is a necessary yet insufficient step towards realising the ideal of deliberative democracy (Felicetti 2018, 810). Yet, we should conceive firms not only as a necessary part of a larger deliberative system but also as deliberative systems themselves (Sabadoz and Singer 2017; Singer 2019). Instead of reducing firms to bureaucratic organisations characterised by an authoritarian model of command and control, this perspective enables us to understand that firms are composed of smaller organisational units which can offer favourable or unfavourable conditions for democratic deliberation. ⁴

As in the larger deliberative system, deliberation within firms might serve epistemic, ethical, and democratic functions (Mansbridge et al. 2012). From a normative perspective, the realisation of all three functions is desirable because each function promotes the democratic legitimacy of decision-making processes “by ensuring reasonably sound decisions in the context of mutual respect among citizens and an inclusive process of collective choice” (Mansbridge et al. 2012, 12). Legitimacy increases the chance that people “will agree, willingly, to the terms of their common cooperation,” which seems particularly valuable in the context of workplace authority (Mansbridge et al. 2012, 12). Yet the three functions might not be equally balanced or might come into conflict with each other. Therefore it is crucial to acknowledge potential imbalances or conflicts when assessing a deliberative setting. For example, firms might be more inclined towards the epistemic function of deliberation due to efficiency constraints (cf. Gerlsbeck and Herzog 2020). Furthermore, I assume that high-quality deliberation indicates that a deliberative system can perform epistemic, ethical, or democratic functions. Later on, I will draw on Dryzek’s (2009) concept of deliberative capacity to assess whether self-organised teams can create spaces for inclusive, authentic, and consequential deliberation within firms (see later discussion). More specifically, I assume that these three criteria can be used to evaluate the quality of deliberations.

The article focuses on self-organised teams for two reasons. Firstly, self-organised teams tend to share decision-making power equally among their members (see Frega 2021). Power sharing is likely to promote the creation of spaces for collective decision-making processes in the form of voting or deliberation. Thus self-organised teams represent a promising starting point to study the potential of democratic deliberation within firms. Secondly, self-organised teams have a direct and immediate impact on the working lives of employees. They shape how employees interact and cooperate on a daily basis, which sets them apart from other organisational subunits of firms (e.g., boards of directors, larger subdivisions). In this regard, self-organised teams have a unique potential to promote the direct participation of employees in deliberative processes within firms and, by doing so, to strengthen the epistemic, ethical, and democratic functions of the firms’ deliberative systems.

Two Approaches: Stakeholder Deliberation and Workplace Deliberation

There are two distinct approaches for making firms more deliberative. First is stakeholder or business deliberation, that is, deliberation among firms or between firms and their stakeholders (Felicetti 2018). Stakeholder deliberation represents a “political” form of corporate social responsibility to promote an authentic, inclusive, and communicative exchange with stakeholders or their representatives (e.g., Palazzo and Scherer 2006; Rasche and Esser 2006; Goodman and Arenas 2015); it is usually organised in “multi-stakeholder initiatives” (e.g., Rasche 2012; Mena and Palazzo 2012; de Bakker, Rasche, and Ponte 2019), the main objective of which is to establish “self-regulatory governing arenas” (Rasche 2012, 679).

Second is “workplace deliberation,” understood as democratic deliberation within firms (Felicetti 2018, 805). To date, only a few scholars have engaged with the topic of workplace deliberation. An early discussion of “worker-self-management” and the “public control of investment” can be found in Cohen ([1989] 2003). According to Cohen, the “economic basis of deliberative democracy” requires “worker-managed firms” because they provide “favourable conditions” for the exercise and development of “deliberative capacities” (46). Even earlier, Gustavsen (1985, 470) made an argument for introducing Habermasian “rational dialogue as the main lever in democratic workplace reform,” an approach that gained attention in Scandinavia (see Gustavsen 1985; Pålshaugen 1998; Brøgger 2010). ⁵

Recently, Felicetti (2018) has reintroduced the topic of workplace deliberation to debates among deliberative democrats and business ethicists. Yet, except for Felicetti, none of these contributions have looked at workplace deliberation from a systemic perspective. Furthermore, none of these contributions have looked at self-organised teamwork as a possible way to institutionalise workplace deliberation, as I propose in this article. Both approaches, that is, workplace deliberation and stakeholder deliberation, are valuable and complementary strategies for making firms more deliberative (Felicetti 2018).

Objections

However, each of the two strategies involves specific challenges. In particular, workplace deliberation faces four objections. Firstly, the control right of shareholders or private owners could lead to situations in which deliberative decisions are overturned or blocked, although this does not rule out the possibility of deliberation within firms per se. In large corporations, shareholders are usually not involved in the daily decision-making processes, which would leave some space for deliberation within firms. Nonetheless, the democratisation of corporations might require some form of legal restriction of the control rights of shareholders (for proposals, see Blair and Stout 1999; Ellerman 1999; Ferreras 2017). ⁶

An empirical example of such models can be found in the German co-determination system in the coal, iron, and steel industry, where joint-stock companies, private limited liability companies, and incorporated companies with more than one thousand employees are obligated by law to guarantee the equal representation of employees and shareholders on the supervisory board (Weiss and Schmidt 2008, 250–52). A second option could be the transformation of capitalist firms into worker cooperatives (Dow 2003, 262). Accordingly, it seems helpful to think about workplace deliberation “along a continuum” of corporate governance models (Felicetti 2018, 807–8). Thus a worker cooperative might be more likely to achieve a higher deliberative capacity than a shareholder-orientated corporation.

Secondly, economic theories of the firm usually assume that a “politicisation” of corporate governance leads to higher transaction costs and losses in efficiency (e.g., Alchian and Demsetz 1972; Jensen and Meckling 1979; Hansmann 1990), potentially putting deliberative firms at a “competitive disadvantage” (Felicetti 2018, 809). However, some studies suggest that democratic firms are more efficient than, or at least on par with, capitalist firms (e.g., Craig and Pancavel 1995; Fakhfakh, Pérotin, and Gago 2012; Pérotin 2016). Overall, the empirical evidence is far from conclusive (Dow 2003, 261–62). But there is certainly a “maximal viability horizon” for democratic norms in firms (Singer 2018, 840). As Singer points out, firms must facilitate economic activities more efficiently than the market—otherwise, they would cease to exist as a form of economic coordination.

Thirdly, workplace deliberation, like other forms of employee participation, might run into paradoxes in which democratic processes produce unethical or undemocratic outcomes. This usually happens when, “in the pursuit of one goal, the pursuit of another competing goal enters the situation … to undermine the first pursuit” (Stohl and Cheney 2001, 354). In most workplaces, this competing goal will be the enhancement of efficiency for the sake of profit maximisation. For example, employees could decide in democratic deliberations that it would increase the efficiency of their team to introduce a peer-based monitoring mechanism to track their working hours, which could lead to an intensification of control through peer pressure. This would run contrary to the ethical function of deliberation. In consideration of the possibility of such “paradoxes of participation,” one might argue that a form of “workplace republicanism” would be better suited to protect the basic interests of employees (Stohl and Cheney 2001, 352; see also Hsieh 2005). But while I agree that workplace deliberation needs to be complemented with republican arrangements, it seems unclear to me why the two models should be seen as mutually exclusive. For instance, most of the firms in the sample of this study implemented deliberative processes while having republican arrangements in the form of the German co-determination system (e.g., work councils, dismissal protection).

Fourthly, the norm of profit seeking might restrict “the range of deliberated-upon topics … in a rather undemocratic fashion” (Sabadoz and Singer 2017, 204). Accordingly, Sabadoz and Singer argue that we should not “mistake deliberation in the service of profit-seeking for democratic deliberation” (204, emphasis original). Yet a growing number of firms pursue not only the objective of profit seeking but also ethical or social welfare objectives (Mitchell et al. 2016). This shows that firms can challenge the norm of profit seeking even in market contexts. However, I agree with Sabadoz and Singer (2017) that this can be challenging. Thus future research on “workplace deliberation” should think about how the norm of profit seeking can be subject to deliberation within firms.

More generally speaking, Sabadoz and Singer (2017) argue that corporations are not a suitable venue for deliberation; they take it that the “social theoretic foundations of deliberative democracy” exist in a hostile relationship to “the corporation’s market context and the nature of why people interact with it,” which must be understood as bargaining rather than deliberation (Sabadoz and Singer 2017, 205–6). This view is similar to theories of functional differentiation which assume that organisations (e.g., firms) function according to a logic that is detached from the moral norms of everyday life (see famously, on “lifeworld” vs. “system,” Habermas [1981] 1987). Thus the market enables human cooperation through instrumental means rather than norms (Sabadoz and Singer 2017, 194). Accordingly, the behaviour of firms is by market forces and profitability calculations. Therefore deliberations within the firm would be limited to choosing the right means to a pre-given goal rather than finding collective ends. Such instrumental deliberation might perform an epistemic function, but it would certainly be restricted in its ability to perform an ethical and democratic function.

However, this view of the firm is based on a too simplistic conception of organisational life. Surely firms are shaped by market forces, but they are also shaped by human beings who are capable of moral agency and norm-orientated behaviour (Herzog 2019). Furthermore, profit maximisation might be the ultimate goal of the firm, but in practice, this goal needs to be further specified or even adjusted. Thus there seems to be enough leeway for democratic deliberations concerning questions of how profits should be made (see Gerlsbeck and Herzog 2020). Deliberative processes in the firm might specify, reshape, or expand upon the goal of profit maximisation and are not necessarily limited to instrumental reasoning. Moreover, Blau (2020, 7) points out that even instrumental rationality is an “essentially contested concept” because “‘good’ or ‘best’ means” are not always obvious or indisputable. A similar argument has been made by Richardson (2002) concerning the role of administrative agencies in the process of policy decision-making: he assumes that agencies are capable of more than just finding the best means to realise the ends set by the legislature because these ends are often ambiguous and need further interpretation. A similar point can be made about firms.

Thus the proposal to see firms as parts of a larger deliberative system and the possibility of deliberation within firms can be defended against these objections. In the next section, I turn to the methodology, on the basis of which I argue that workplace deliberation is indeed a promising way forward for deliberative democrats.

METHODOLOGY

The study is based on in-depth interviews with employees in firms which practice various forms of self-organised teamwork. ⁷ The interviews were semi-structured and included a combination of open-ended, probing, and background questions. Each interview addressed a broad range of issues related to the employees’ experiences of working in a self-organised team. The list of questions was clustered around issues like corporate culture, decision-making processes, and personal motivation. The selection of issues was guided by studies in organisational theory, business ethics, and deliberative democratic theory.

As Ercan, Hendriks, and Boswell (2017, 197) point out, such interpretative methods are particularly good at giving us a context-sensitive understanding of the perspectives of the involved participants. Moreover, the study of deliberation in firms is still in an exploratory stage, which lends support to an interpretative approach (Felicetti 2018, 812).

However, interpretive research is confronted with challenges to guarantee its validity and reliability (Randall and Gibson 1990; Liedtka 1992). In particular, an interview-based approach faces the risk of social desirability bias (Nederhof 1985; Randall and Fernandes 1991). I have tried to counter social desirability biases by assuring the anonymity of the participants and by emphasising that there are no wrong or right answers. Also, “presuming” questions, which would already imply a certain notion of social desirability, were avoided.

Materials and Sample-Strategy

I conducted sixteen employee interviews in six German firms. ⁸ In total, nine contact approaches were made, with a response rate of 66 per cent—a high rate, given the entry problems of empirical research into business ethics (Randall and Gibson 1990, 464). The fact that my sample only consists of German firms might raise the question of whether this introduced a country-specific bias, as German firms operate in a coordinated market economy with co-determination (Hall and Soskice 2001; Page 2011). Thus it is an open question whether similar findings could be observed in firms that operate in liberal market economies. Although it is conceivable that the phenomena the interviewees described could also have taken place outside Germany, further research on this question is needed.

The firms varied significantly in their numbers of employees (from fourteen to more than three hundred thousand). They came from the industries transport and logistics, hospitality, urban planning, consumer goods, engineering/technology, and consulting. The urban planning firm was organised as a worker cooperative; the others had traditional corporate governance structures. I used a sampling strategy under which “information-rich cases” (Patton [1990] 2002, 169) were selected, first through media coverage about self-organised work, and then through a snowballing method. I terminated sampling when the information I received from the interviewees showed signs of saturation (Sandelowski 2008, 875).

Yet this sampling strategy creates limitations when it comes to the external validity of the findings. In particular, the combination of relatively small sample size and non-random selection allows no statistically representative conclusions. However, the aim was not to make generalisations but rather to gain in-depth insights into specific cases that can enrich our normative theorising (Herzog and Zacka 2019). Thus the results of the study should be viewed as suggestive rather than conclusive. I tried to interview more than one employee from each firm, which was successful in five out of six cases. When possible, I contacted interviewees personally to limit the impact of self-selection bias (James 2006). Each interview was conducted in person in 2019. The interviews were tape recorded, with the interviewees’ consent, and transliterated.

The average age of the interviewees was forty-two years, and they had worked on average for thirteen years in their respective professions. I interviewed eight men and eight women. Their organisational positions varied from entry level to CEO, although most held middle or higher management positions. Thus managers and white-collar workers are somewhat over-represented, so the results should be interpreted with caution when it comes to teams that involve greater numbers of low-skilled workers. Yet, for certain results, it is plausible to draw conclusions that go beyond teams of white-collar workers. Furthermore, the studied self-organised teams were highly diverse with regard to the cultural and educational backgrounds of the team members. The topics on the agendas of the teams’ deliberations included daily administrative and operational decisions, for example, concerning resource allocation, product development, or staff-related issues. In the smaller firms, the teams’ deliberations also included strategic decisions concerning the larger organisational structure and the firm’s long-term mission.

Data Analysis

The transcripts were subjected to a qualitative content analysis that involved four systematic steps (Mayring 2015). Firstly, I defined the textual corpus and the units of analysis. Secondly, I deduced thematic categories from the research question and previous readings. These categories included 1) corporate culture and team spirit, 2) decision-making processes and deliberation, 3) employee motivation in self-organised teams, and 4) obstacles for deliberation in self-organised teams. I analysed the material on the basis of each category, and I assigned relevant text passages a code. Thirdly, I reduced the relevant text passages to their thematic core and adjusted them to a similar level of abstraction. Building upon this reduction process, I looked for patterns in the material that could be converted into general themes. Fourthly, I assigned these themes a code and matched them with exemplary text passages, for a final cross-check.

DELIBERATION IN SELF-ORGANISED TEAMS

Self-Organised Teamwork

In the literature on teamwork, one can find various definitions of self-organised teams (e.g., Hackman 1986; Manz 1992). I will rely here on the definition by Cohen and Bailey (1997, 241), who describe a team as “a collection of individuals who are interdependent in their tasks, who share responsibility for outcomes, who see themselves and who are seen by others as an intact social entity embedded in one or more larger social systems…, and who manage their relationships across organizational boundaries.” The most common criterion to distinguish different types of teams is the level of the team’s discretion or autonomy (Hackman 1986; Manz 1992). One can think of a “continuum” between self-control and external control. For example, Manz (1992, 1129) distinguishes between four types of teams: “externally managed teams” with the lowest level of self-control, followed by “participative teams” (e.g., quality circles), “self-managed teams” (with limited autonomy over the direction or purpose of the work), and “self-leading teams” (to which he also refers as “self-governing teams”). In my sample, one can find “self-managed teams” as well as “self-leading teams.” I therefore use the term self-organised team to describe a team that can be either semi-autonomous or autonomous.

The origins of self-organised teamwork can be found in theories of employee participation and the human relations approach in psychology, pioneered by Kurt Lewin and Elton Mayo (Frega 2021, 8). These approaches developed a holistic perspective on work, as influenced not only by psychological (e.g., needs, desires, goals) but also by social circumstances (e.g., norms, identity, membership) (Bruce 2006, 180). Relatedly, they focussed on “groups, group standards, and group decision making and … nonpecuniary as opposed to pecuniary rewards” (Strauss 2006, 780). However, the human relations school did not focus on the implementation of formal institutions for employee participation (Strauss 2006, 780). A stronger connection to formal employee participation was formed in the 1960s and 1970s, under the label of “job redesign” (Strauss 2006). The work by Abraham Maslow and Douglas McGregor introduced a focus on human needs and motivation. Thus autonomy and meaningfulness were considered as determinants of efficiency, employee motivation, and the satisfaction of post-material needs (e.g., Hackman and Lawler 1971; Hackman and Oldham 1976). Firms began to search for alternatives to bureaucratic decision-making, divisional fragmentation, and hierarchical chains of command. As a result, employee participation and participatory management techniques gained attention, together with “job enlargement, job enrichment, and work teams” (Strauss 2006, 781). These changes had a stronger political dimension because they involved an element of “power equalisation” (Strauss 2006, 781). On a scholarly level, this shift was accompanied by an interest in workplace democracy, for example, worker cooperatives or Scandinavian experiments with autonomous work groups (Pateman 1970; Emery and Thorsrud 1976).

In the 1980s and 1990s, scholars and practitioners continued to experiment with employee participation. Teamwork became the most common instrument to promote employee empowerment (Kuhlmann 2002; Pruijt 2003). Moreover, there was a managerial impulse to introduce organisational changes, triggered in part by fears of being left behind by “lean” Japanese production methods, with their focus on waste minimisation (Strauss 2006, 782; see also Womack, Jones, and Roos 1990). A common feature of these approaches was a focus on decentralised decision-making, which was institutionalised in the form of autonomous or semi-autonomous groups and higher levels of responsibility for employees (Streeck 1987, 321). However, scholars have criticised that these developments could lead to work intensification; stricter work rules; and overly demanding performance standards, problematic forms of peer pressure, and even health risks (e.g., Berggren 1993; Lewchuk and Robertson 1995; Landsbergis, Cahill, and Schnall 1999; Pruijt 2003). Furthermore, some scholars have argued that employee empowerment is restricted by the operative objectives of the firm and represents only a situational suspension of hierarchical structures (Babson 1995).

Despite this criticism, other industries followed the example of the automobile industry and began to experiment with similar organisational models (Holweg 2007, 431). In Europe “anti-Tayloristic reforms” like “socio-technical systems design, industrial democracy, and humanisation of working life” spread further (Pruijt 2003, 79), a trend fuelled by consulting companies and human resources managers. However, it reached its peak by the end of the 1990s (Batt 2004). Employee survey data show that the number of self-managed teams stagnated in the United States: 40.5 per cent in 1992 compared to 41.1 per cent in 1997 (Osterman 2000, 186). The reasons for this stagnation seem to have been a pushback by supervisors and middle managers, a lack of institutional protection, and shareholder-orientated reforms of corporate governance (Boes et al. 2018, 187). These factors outweighed the positive effects of teamwork and employee participation on performance and job satisfaction (e.g., Cotton 1993; Cohen and Bailey 1997; Wood and de Menezes 2011). As a result, most experiments with employee participation and self-organised teams never progressed beyond the initial stage of experimentation (Strauss 2006, 783; see also Dörre 1996).

However, the digital transformation of work and the rising volatility of globalised markets have revived the interest in teamwork and employee empowerment (see Malone, Laubacher, and Morton 2003; Schrauzer 2016; Boes et al. 2018). More specifically, companies have begun to search for more agile organisational models to handle continuous customer involvement and the demands of short-cycle product development (Boes et al. 2018; Burchardt and Maisch 2019). Often, this search led to the replacement of hierarchical structures with self-organised teams to improve knowledge sharing and cross-functional collaboration (Schwarzmüller et al. 2018). In particular, the principles of “agile” software development gained attention among companies (Boes et al. 2018). While originally confined to the context of software development, “agility” has developed into a broader organisational framework that is used in various industries (Rigby, Sutherland, and Takeuchi 2016).

In contrast to earlier forms of teamwork, the focus of this process has been the management level and not the shop floor, which might partly explain why the connection to traditional themes of the labour movement (e.g., industrial democracy, worker rights) is much weaker or even absent in these organisational models (Boes et al. 2018, 184; see also Singe and Tietel 2019, 252). Instead, there is a stronger emphasis on individual autonomy, flexibility, and the activation of innovation capabilities (Kalff 2019; Singe and Tietel 2019).

Today, the most common agile method is the scrum framework (Sutherland and Schwaber 2013). Scrum is a project management approach that builds on short cycles of two or four weeks (“sprints”). A self-organised and cross-functional team is formed and creates a “backlog,” which contains the central customer demands as a road map for the project. From each cycle to the next, the team defines which tasks are to be executed, when, and how; after each cycle, a “usable” product or solution must be presented. Accordingly, the project manager is replaced by an empowered and self-organised team. Scrum introduces two functional roles, namely the “scrum master,” who functions as a process facilitator, and the “product owner,” who represents the perspective of the customer (Boes et al. 2018, 189), but these roles do not hold a specific authority over other team members.

To sum up, agile methods like scrum grant employees a significant level of discretion by dismantling the hierarchies of command and control and replacing them with self-organised team structures. In the sample, one firm adopted a scrum framework in several of its subdivisions, and another company took inspiration from agile methods to restructure its human resources department.

A second approach that contributed to the renaissance of self-organised teams is holocracy, which also originated in the context of software development and borrows central elements from agile methods and lean production. It replaces top-down hierarchies with self-organised teams, the “circles,” which can emerge and change over time (van de Kamp 2014, 16). There is a predefined function for each circle and a hierarchy of self-organised circles consisting of sub-circles within larger circles (Bernstein et al. 2016). Furthermore, holocracy replaces job titles with roles with clear functional purposes within the self-organised teams. Thus holocracy distributes authority widely throughout the organisation by empowering teams and individuals in their specific roles.

Another element of holocracy is that employees ratify a “constitution” that defines the general rules for creating, changing, and dissolving circles (Bernstein et al. 2016). While agile methods and lean production are implemented to reorganise specific parts of a company, holocracy represents a framework for a fully self-managed organisation, in which the employees share authority over the goals of the firm, accountability for the operations, and discretion over resources and information use (Bernstein et al. 2016). My sample includes two holocratic firms and one company that adopted elements of holocracy.

Thirdly, there has been renewed interest in lean production. Many companies have begun to combine digital technologies with lean methods to increase the flexibility of their manufacturing processes and to reduce waste (see Burggräf et al. 2017; Meissner et al. 2018). Accordingly, lean practices like self-organised teams also regained attention among companies (Martinez-Jurado et al. 2014).

Overall, agile or lean models and holocracy represent alternatives to top-down hierarchies and bureaucratic control. They are particularly important for firms that need to adapt quickly to a changing market environment, and it is likely that with the digital transformation, more firms will feel such pressures. Although these market pressures are the driving force behind the interest in these organisational models, this development might also hold a democratic potential by creating spaces for decentralised decision-making, self-organisation, and employee empowerment within firms. As Slater and Bennis (1990, 169, emphasis original) put it, “democracy becomes a functional necessity whenever a social system is competing for survival under conditions of chronic change.” Thus firms must organise themselves in more democratic ways to become more agile, flexible, or lean (see Jones 1995; Safari, Salehzadeh, and Ghaziasgar 2018; Singe and Tietel 2019).

However, self-organised teamwork contains not only a democratic but also an exploitative potential that can lead to work intensification, peer pressure, and heteronomous work (see Kelly 1995; Hodgson and Briand 2013). Thus there is an ambivalence in these organisational frameworks—on one hand, they have the potential to create democratic spaces within firms, but on the other hand, these spaces are created to improve organisational efficiency and gain market advantages.

According to Yeoman (2014, 155), participation at the workplace becomes fully democratic if it “implies a transformation of the authority structure in which each employee has equal decision-making power.” Yet, the transformation of authority structures depends on the specific organisational design of the team setting and its interaction with other parts of the organisation (Boes et al. 2018).

Assessing the Deliberative Capacity of Self-Organised Teams

To assess whether deliberation within firms can perform an epistemic, ethical, or democratic function, it is necessary to introduce “a template to evaluate the conditions that support the various functions of good deliberation” (Mansbridge et al. 2012, 13). Despite the growing number of empirical studies on democratic deliberation, there is a “lack of agreement on a uniform definition of deliberation from which reliable empirical measures can be derived” (Black et al. 2011, 338). ⁹ Accordingly, scholars have tried to define what good deliberation is and the criteria on the basis of which it can be operationalised. I draw on Dryzek’s (2009) concept of deliberative capacity, for three reasons. Firstly, this concept is based on a systemic perspective of deliberation and therefore shares a central premise of this article (Dryzek 2009, 1384–88). Secondly, it is compatible with a qualitative approach to the study of deliberation (see Stevenson and Dryzek 2012; Felicetti, Niemeyer, and Curato 2015). Thirdly, it is rooted in a comprehensive approach to the democratisation of society and therefore compatible with workplace democracy. Dryzek (2009, 1382) defines deliberative capacity as “the extent to which a political system possesses structures to host deliberation that is authentic, inclusive, and consequential.” The concept thus introduces three criteria: authenticity, inclusiveness, and consequentiality. I assume that high levels of authenticity, inclusiveness, and consequentiality indicate that a deliberative system shows signs of high-quality deliberation and can perform epistemic, ethical, or democratic functions.

Authenticity

Authenticity is given if deliberations “induce reflection noncoercively, connect claims to more general principles, and exhibit reciprocity” (Dryzek 2009, 1382). On a theoretical level, this definition seems to be rather uncontroversial (Felicetti, Niemeyer, and Curato 2015). However, empirically, it can be challenging to evaluate whether deliberation is authentic. In particular, it seems almost impossible to make “a judgment about a person’s true vs stated preferences” (Steenbergen et al. 2003, 26). Thus it is necessary to further clarify Dryzek’s criteria for authentic deliberation.

Firstly, non-coercion is given if a situation is free from deception (e.g., lying), strategic manipulation (e.g., withholding information, misinterpreting facts), domination, subordination, the use of force, and threats of sanctions (Warren 2006; Mansbridge et al. 2010). ¹⁰ Secondly, deliberative processes “induce reflection” if they lead participants to revise their preferences or views in light of discussions, produce informed and reasonable opinions among participants, and consider the arguments or viewpoints of others (Felicetti 2016, 12–13). Thirdly, deliberative processes are authentic if they advance claims which are connected to universal principles or appeal to some “overarching interest” (Chambers 2003, 309). One could also speak of a “public-spirited perspective” that should be reflected in the arguments (Gutmann and Thompson 2004, 10). In the context of the workplace, such a perspective might include the appeal not only to broader political or social values but also to corporate interests if we consider firms as actors that produce socially desirable goods or services and provide employment opportunities. Fourthly, deliberative processes are authentic if they exhibit reciprocity. Dryzek (2009, 1381) defines reciprocity as a “deliberative virtue” which is “stated as communicating in terms that others can accept.” Yet, reciprocity involves not only how we communicate with each other but also the form and content of the reasons we give to each other. More specifically, we need to provide reasons in terms that others can reasonably accept (see Habermas 1996; Rawls 1993). Accordingly, reciprocity demands reasoning that is “mutually acceptable and generally accessible” (Gutmann and Thompson 2004, 7).

Inclusiveness

Inclusiveness refers to “the range of interests and discourses present in a political setting”; it differentiates democratic deliberation from other forms (Dryzek 2009, 1382). According to Dryzek (2010, 10), inclusive “deliberation requires the opportunity and ability of all affected actors (or their representatives) to participate.” This definition of inclusive deliberation resembles the “all affected” principle (e.g., Cohen [1989] 2003; Goodin 2007). Yet, taken at face value, this principle would imply the inclusion of potentially every actor that is somehow affected by the deliberative process (Goodin 2007). Although this might be desirable from a normative perspective, it is unfeasible in empirical terms. Therefore it is necessary to further specify the “all affected” principle. To do so, I draw on Fung’s (2013, 247) formulation of the principle, which states that “an individual should be able to influence an organization if and only if that organization makes decisions that regularly or deeply affect that individual’s important interests.” Although the “regularly or deeply affected” formulation offers a more specified standard regarding who should participate in a deliberative process, it still begs the question of what are “important interests” and what is meant by “being affected.” According to Zimmermann (2017, 3), interests are important or “morally significant” if they affect the autonomy and well-being of an actor. Furthermore, I assume that actors are affected in a way that justifies granting them participation rights if they are affected by power relations that involve actual or potential coercion. To summarise these points, I will consider a deliberative process as inclusive if it gives all actors who are regularly or deeply affected in their interest in autonomy and well-being the opportunity and ability to participate.

Whereas Dryzek (2009) defines inclusivity in terms of interests and discourses, it seems helpful also to look at the effects of social inequalities and demographic factors when it comes to the inclusiveness of deliberative processes (Felicetti 2016, 16; see similarly Fraser 1990; Young 2000). Furthermore, communicative or cultural standards should be considered concerning the question of who is included or excluded in a political setting (cf. Sanders 1997; Young 2000). To put this in broader terms, inclusiveness should embody “a norm of moral respect” and “political equality” (Young 2000, 23; see similarly Kuyper 2015, 323). Thus interactions must be characterised by mutual respect and provide participants with an “effective opportunity to express their interests and concerns” (Young 2000, 23; cf. Mansbridge et al. 2010; see also Moscrop and Warren 2016 on agenda-setting processes).

Consequentiality

Deliberative processes are consequential if they “have an impact on collective decisions or social outcomes” (Dryzek 2009, 1382). However, this impact does not have to be “direct,” which means that “deliberation need not involve the actual making of policy decisions” (Dryzek 2009, 1382). Kuyper (2015, 324) summarises some of the potential outcomes of a consequential deliberative process as “influence on decision-makers, change to institutional and cultural rules, and reflexive alterations of different discourses and interests.” Furthermore, Felicetti (2016) makes three important points when it comes to the consequentiality of deliberations. Firstly, it is important to acknowledge that the consequences of deliberations sometimes “unfold over time”; secondly, consequentiality is context-specific; thirdly, we can only assess deliberative consequentiality if there is authentic and inclusive deliberation in the first place (Felicetti 2016, 19). Otherwise, we would only assess “non-deliberative consequences” which are the result of “non-inclusive and non-authentic deliberation” (Kuyper 2015, 324). Thus deliberative consequentiality “depends on authenticity and inclusivity” (Felicetti 2016, 18).

RESULTS

In this section, I analyse employees’ assessments of group discussions and collective decision-making processes within self-organised teams. I focus on the question of whether these processes showed evidence of authentic, inclusive, and consequential deliberation. Furthermore, I discuss whether the structural conditions of the team setting enable or restrict authentic, inclusive, and consequential deliberation (see Felicetti 2016; Holdo 2020).

Is There Authentic Deliberation in Self-Organised Teams?

The first criterion for authentic deliberation is the non-coercive character of interactions and communicative processes. As described earlier, most workplaces fall short of this condition and are subject to “managerial authoritarianism and domination” (Breen 2017, 1). However, self-organised teamwork opens up a space for non-coercive deliberation by suspending managerial authority. I base this claim on three observations that emerged in the interviews. Firstly, all the teams internally abolished management positions and replaced them with functional roles. These roles do not grant general discretionary power over other team members; instead, their discretionary power was limited to their functional domain, or, as one employee from a holocratically organised company put it, “that is the principle of shared authority. Yes, each role has its domain, discretionary rights so to speak, but the team determines who holds which role, or whether a role exists at all.” ¹¹ The teams replaced managerial authority by functionally limited and democratically legitimised roles which reduced coercion based on subordination, threats, or sanctions. However, employees in two teams described the problem that functional roles can, over time, transform into informal hierarchies. As one interviewee remarked, “We then started, maybe to be more democratic, to distribute tasks to [employee] tandems.… In the end, you could notice that hidden hierarchies emerged again.” The persistence of these “hidden hierarchies” might be explained by the fact that certain members of the teams developed expertise on which the others depended. Thus they might have contributed to the epistemic function of the teams’ deliberations. Nonetheless, all interviewees felt that there was less hierarchy than in traditionally managed teams.

A second aspect that contributed to the absence of coercion was the equal and open access to the information within teams. ¹² Hierarchical organisations are prone to conflicts that can lead to the strategic use of information (Herzog 2018, 110). Yet, by replacing top-down hierarchies with lateral relationships, the teams were able to minimise this problem. As a human resources manager put it, “So we are all equally informed. Before, there was less information, and you had the feeling that an agenda was pursued, but now, topics are transparent and always visible.”

However, employees from the two larger firms reported that they still experienced subordination or domination, particularly when it came to interactions with the firms’ top management. A manager from a transport and logistics company described these interactions as follows: “This lack of ability to be heard with my arguments can be unpleasant from time to time, I have the feeling of not being seen—that bothers me!” Thus interactions with the larger organisational hierarchy can lead to moments of coercion. Another aspect that increased the level of coercion within one team was the instrumentalisation of majority votes. As a manager from a technology company described, “Those majority votes, I think you use this method to make it faster, it is then super-quick and you maybe use it intentionally or unintentionally to steamroller someone when you know that you don’t want to discuss something any longer.” Although voting can be a “democratic necessity of some form[,]… when interests conflict irreconcilably, negotiation to agreement is impossible, or an assembly simply runs out of time,” its instrumental use to cut off deliberations is a coercive violation of authentic deliberation (Mansbridge et al. 2010, 85). However, only one employee from the sample perceived the instrumentalisation of majority voting as problematic. Overall, deliberations within the teams showed very low levels of coercion.

The second criterion that needs to be assessed in relation to authentic deliberation is the question of whether deliberation induced reflections among the participants. The first indicator for reflection would be that team members revised their opinions or preferences in light of the discussions. Interviewees from three teams explicitly reported that they participated in discussions that had led to revisions of their preferences. For example, a manager from a transport and logistics company said, “That is happening every day, honestly! We hear the arguments and the truth is somewhere in between.… Often I become convinced, and sometimes I convince others.” The second indicator of reflections induced by discussion would be that participants consider others’ viewpoints. All sixteen interviewees maintained that they could influence decisions within their teams by bringing forward arguments, which means that they were considered by others. This is aptly captured by a quote from a manager of a transport and logistics company: “Definitely! The decisions we take together are based on the arguments of everyone.” Furthermore, several interviewees said that they experienced group discussions as fair and open.

The third indicator for reflection would be that the deliberations produced reasonable and informed opinions. Again, there is strong evidence for this indicator. Employees from four teams said that the epistemic quality of decisions had improved by becoming self-organised. The interviewees referred to aspects like a broader interdisciplinary exchange, the wisdom of crowds, and better access to information, which resulted in a higher epistemic quality of decisions. The director of executive compensation from a technology company said, “I am convinced after my experience within the self-organisation … that we took more well-informed decisions.” However, the interviewees from a consulting company and consumer goods firm were unsure about the higher epistemic quality of decisions. But they stressed that the legitimacy of decisions had improved by replacing managerial authority with collective decision-making. To sum up, deliberations within the teams produced reasonable and informed opinions.

The third criterion that needs to be assessed in relation to authentic deliberation is the question of whether arguments and justifications appeal to universal principles or some form of overarching interest. There is tentative evidence for this criterion. Employees from all teams maintained that they felt a strong sense of community within their teams or firms. This might be interpreted as an indirect indicator of the presence of shared, overarching interests. Furthermore, six interviewees from three different teams thought that most employees considered the well-being of the company in their decisions and actions. As a project manager from an urban planning firm remarked, “I am so naive to believe that the majority of employees put the well-being of the company first.” This appeal to corporate interests or well-being should be interpreted as a sign of authentic deliberation. Yet, one manager also pointed out that the self-interests of some team members were difficult to combine with the interests of the team: “you asked if people pursue their interests or focus on the company, and there is still a large gap.” Another employee pointed out that decisions which concern the money or the private life of employees were particularly connected to self-interest and therefore difficult to discuss. Thus the evidence is mixed on this point.

The fourth criterion is the question of whether deliberations among participants exhibit reciprocity, for which there seems to be strong evidence. Employees from all firms said that they experienced the discussion culture within their teams as fair, open, and comprehensible. This implies that communicative processes were mutually accepted. The fact that all the interviewees said that they could influence deliberations by bringing forward arguments supports this assumption. As a manager from a consumer goods company remarked, “We have a [team] culture where you talk a lot and conflicts are openly addressed.” Furthermore, employees recognised others’ claims even if they believed that they were wrong. This seems to indicate that employees expressed their views or arguments in terms that others could reasonably accept. Thus not only were the communication processes mutually accepted but so were the form and content of the expressed reasons, which represents another key element of reciprocity (Gutmann and Thompson 1996).

To summarise these findings, there is strong evidence for the three criteria of authentic deliberation in these self-organised teams, namely non-coercion, reflection induced by discussion, and reciprocity. The evidence for arguments and justifications that appealed to universal principles or overarching interests is less strong. Another caveat of these findings is that interviewees might have been subject to social desirability bias because the criteria of authentic deliberation are usually considered socially desirable. Overall, the evidence for authentic deliberation is an indicator of the ability of the studied teams to perform an ethical function by promoting mutual respect among employees and recognising them as “self-authoring sources of reasons and claims” (Mansbridge et al. 2012, 11). Furthermore, the evidence also indicates that the studied teams were able to perform an epistemic function by producing reasonable and informed opinions. However, although the development of informal hierarchies in some teams may have contributed to this epistemic function, it also negatively affected the ability of the teams to perform a democratic function by undermining the equality of the team members.

Is There Inclusive Deliberation in Self-Organised Teams?

Deliberation is inclusive if all actors who are regularly or deeply affected in their interest in autonomy and well-being have the opportunity and ability to participate in the deliberative process. If one applies this definition of inclusive deliberation to a corporate context, it becomes apparent that the power of a company regularly and deeply affects the autonomy and well-being of employees. Yet, sometimes, also other stakeholders, for example, contract workers or suppliers, can be affected in a way that falls under the described definition of inclusiveness and should therefore have the opportunity and ability to participate in the deliberative processes within the firm (see similarly Felicetti 2018). ¹³

Almost all self-organised teams implemented formalised rules or rights that guaranteed that all team members could advance their interests in deliberative processes. ¹⁴ Only one team had no formalised rules concerning collective decision-making processes. In all other teams, a “constitution” or catalogue of rules and rights laid out the exact processes of collective decision-making. The managing director of a consumer goods company aptly captured this aspect: “everybody is subject to the same rules, everybody has the same opportunities, each claim must be justified.” Furthermore, employees in these teams had an equal right to amend changes to the “constitution” of the team.

However, in most cases, the equal right to participate did not include employees who were not part of the team but were still affected by the decisions of the team. Furthermore, the teams did not include contract workers or suppliers in their deliberative processes, which reduced the inclusiveness of their deliberations. Only one team explicitly opened its deliberative processes to other employees. As a manager from this team remarked, “I think that we got a different kind of openness. I notice that the fear to write to us, to approach us, or to tell us things is much less than in other departments. For example, we offer coaching sessions where a part of our group participates, and we offer open spaces to employees from outside.”

However, an equal right to participate might be not enough to guarantee the inclusiveness of deliberation and might not lead to the actual inclusion of all regularly and deeply affected interests. It is therefore important to guarantee that everyone who is affected by a deliberative process has the “effective opportunity to express their interests and concerns” (Young 2000, 23, emphasis added). Such an opportunity can be influenced by aspects like communicative and cultural standards, demographic characteristics, disabilities, character traits, cognitive abilities, and social inequalities. Interestingly, some of the teams introduced mechanisms for reducing the negative influence of these aspects on inclusiveness. Four teams made it mandatory that team meetings include a moderator and sometimes also a time keeper. These four teams also had detailed rules for the workflow of team meetings and collective decision-making processes.

Several interviewees expressed concerns about the difficulties of equally including everyone affected. One manager said that seniority negatively affected deliberations because younger employees felt intimidated and contributed less. Other problems included group size, time constraints, and different levels of knowledge about decision-making rules. The managing director from a consumer goods firm summarised this problem: “It is decisive under what kind of preconditions you go into the system. Employees are not equal, some can speak good German, others can’t speak as well; some are good with numbers, others are not; some people are educated very well or have an instinct for power and can pick up topics or push them through. I think we won’t get far as long as our systems are blind to these aspects.” The quote illustrates that some teams were aware of these problems but still had a hard time addressing them.

The final point to consider is the problem that employees did not always exercise their rights, which is a problem from the perspective of inclusive deliberation. Several interviewees pointed out that this problem was much more pronounced for employees who were used to working in top-down hierarchies. The spa manager from a hotel chain put it like this: “We’ve got flat hierarchies here, however, we have somebody who worked in the army for a long time.… It can be difficult to get somebody like that on board as he does not feel comfortable because he has been shaped differently.” This feeling of uncomfortableness might be connected to the fact that a participative environment combined with high levels of autonomy can be challenging for employees with “a low need for autonomy” and insufficient “self-managing abilities” (Manz 1992, 1133).

However, concerning these problems, one employee remarked that she experienced democracy as a learning process: “So I would translate it [democracy] with participation, cooperation but I also learned in the corporate context that democracy needs to be learned first.” The quote reminds one of Pateman’s (1970, 43) arguments that workplace democracy could provide “an education in the management of collective affairs that is difficult to parallel elsewhere.” Thus the willingness of employees to participate in deliberations might require a longer learning process which might be more successful under conditions of mutual trust and transparency (Boes et al. 2018, 203).

To summarise these findings, there is evidence for inclusive deliberation in all self-organised teams. However, self-organised teams, like other deliberative or empowered spaces, face certain challenges to effectively include all affected interests. Some teams achieved higher levels of inclusiveness than others. In particular, the four teams which relied on moderators and highly structured team meetings had a higher degree of inclusiveness than the other two teams. This evidence corresponds with previous studies on small-group deliberation (e.g., Fishkin and Luskin 2005; Grönlund, Herne, and Setälä 2015). The evidence for inclusive deliberation is an indicator of the ability of the studied teams to perform a democratic function by including multiple interests, concerns, and claims equally. Furthermore, the inclusion of multiple perspectives also indicates that the deliberative processes performed an epistemic function by bringing together different perspectives. This is supported by the fact that several employees described a higher epistemic quality of decisions compared to the time before becoming self-organised.

Is There Consequential Deliberation in Self-Organised Teams?

Deliberative processes are consequential if they “have an impact on collective decisions or social outcomes” (Dryzek 2009, 1382). So, for my study, the question is whether the deliberations within the self-organised teams had an impact on collective decisions or social outcomes. On the basis of the interview material, the answer is a clear yes: in all teams, deliberations had an impact on collective decisions or social outcomes. Although the teams also used aggregative mechanisms, such as voting, they relied mainly on deliberative processes to make collective decisions. Employees from all teams reported that they felt that their arguments had an impact on collective decisions. As one manager put it, “If somebody makes a decision memo and presents it in the general committee, then my co-workers or I can change this decision by presenting our evidence.” Furthermore, most collective decisions had a direct impact on the organisational life, as can be illustrated by an example concerning the dismissal of an employee. During his probation period, this employee showed evidence of disloyal behaviour. Even after several meetings to resolve the problem, his behaviour did not change. Confronted with this problem, the team initiated a deliberative process to decide collectively whether the employee should be dismissed. An employee from the firm described this process in the following way: “Actually, we dismiss [employees] as a team and these decisions are not taken by the executive board.” Furthermore, he emphasised that the decision to dismiss the employee was primarily based on the intention to protect the organisational integrity of the team: “There are things where we have to say ‘sorry, that is not okay.’ This does not fit with how we want to work together, and if the person does not change those things after two discussions, we start a process in which we collectively decide to let somebody go or not.” Thus efficiency considerations played only a secondary role in the deliberative process.

However, not all teams had the same level of consequentiality related to their deliberations. Whereas some teams were autonomous in their decisions, others can only be considered semi-autonomous. Thus the top management could interfere with collective decisions or even block them, which decreased the level of deliberative consequentiality (cf. Sabadoz and Singer 2017). To summarise these findings, there is strong evidence for consequential deliberations in all self-organised teams. However, the level of consequentiality varied between teams, depending on their status in the larger organisational hierarchy of the firm. The evidence of consequential deliberation is an indicator of the ability of the studied teams to perform an ethical function by treating employees as autonomous agents who take part in the collective decision-making processes of their workplaces.

CONCLUSION

In this article, I have argued that deliberative democrats should consider firms as parts of a larger deliberative system. Following Felicetti (2018), I have identified two complementary strategies to make firms more deliberative and focussed on one of them, namely workplace deliberation. Furthermore, I adopted the view that we should think of firms as deliberative systems in themselves (Sabadoz and Singer 2017; Singer 2019). This perspective allows us to understand how smaller organisational units of firms can exhibit a high deliberative capacity even while other parts remain non-deliberative. I have identified self-organised teams as organisational units within firms that exhibit a high level of deliberative capacity and can create spaces for democratic deliberation. Furthermore, this high level of deliberative capacity indicates that the studied teams were able to perform an epistemic, ethical, and democratic function. I have based this claim on sixteen in-depth interviews within six German firms that practice various forms of self-organised teamwork. The studied teams showed evidence of authentic, inclusive, and consequential deliberation. The level of inclusiveness and consequentiality varied between teams. Concerning inclusiveness, this variation might be explained by the absence of highly structured and moderated team meetings. Consequentiality seems to depend, to a large extent, on the status of the team in the larger organisational hierarchy of the firm. Thus autonomous teams exhibited deliberative processes that were more consequential than those of semi-autonomous teams. However, these findings should be viewed as suggestive rather than conclusive evidence for the deliberative capacity of self-organised teams due to the relatively small sample size of the study.

The comprehensive democratisation of firms would require additional venues for democratic deliberation. Deliberative democrats should look for additional spaces within firms that could (or should) offer favourable conditions for deliberation. In addition to issues that concern their daily working lives (Felicetti 2018, 806), workers should be able to participate in deliberations that concern decisions at the level of the firm. Managers and employees might think differently about democratic reforms in the workplace. In particular, middle managers and supervisors are likely to reject the use of self-organised teams to protect their status and discretion (Batt 2004). To overcome such resistance to democratic reforms, it is crucial that workers also participate in decision-making processes at the level of the firm (e.g., board-level participation). Moreover, workers should be represented in deliberations that go beyond single firms and, by doing so, have an influence on their industrial sector or the society as a whole (see similarly Hussain 2009). This has been historically realised in the form of trade unionism (Webb and Webb 1897). At first sight, unions’ collective bargaining seems to be opposed to deliberation. However, it might involve moments of “deliberative negotiation” (Mansbridge et al. 2010) if the involved actors pursue their “self-regarding conflicting interests…, on fair terms and with mutual respect,” and aim for a “fair compromise” (Naurin and Reh 2018, 728 and 732). Thus collective bargaining might qualify as a space for deliberation beyond the single firm.

To sum up, a truly democratic firm should guarantee that workers can deliberate on decisions that affect their working lives at all three levels (work unit, firm, and industrial sector/society). This three-level approach to workplace deliberation would contribute to the three functions of the larger deliberative system (Mansbridge et al. 2012). However, further empirical research is needed to assess the deliberative capacity of all three spaces and to better understand their role in the larger deliberative system.