When democratic principles are not enough: Tensions and temporalities of dialogic stakeholder engagement Abstract Stakeholder engagement and dialogue have a central role in defining the relations between organisations and their internal and external interlocutors. Drawing upon the analysis of dialogic motifs, power-conflict dynamics and socio-political perspectives, and based on a set of interviews with the stakeholders of a consumer-owned co-operative, the research explores the dialogic potential of stakeholder engagement. The analysis revealed a fragmented picture where the co-design and co-implementation aspects

Stakeholder engagement and dialogue have a central role in defining the relations between organisations and their internal and external interlocutors. Drawing upon the analysis of dialogic motifs, power-conflict dynamics and socio-political perspectives, and based on a set of interviews with the stakeholders of a consumer-owned co-operative, the research explores the dialogic potential of stakeholder engagement. The analysis revealed a fragmented picture where the co-design and co-implementation aspects were mainly related to the non-business areas of co-operative life, while business logic dominated the most central aspects. Stakeholder engagement was mainly related to consensus building, while dialogic engagement based on a pluralistic understanding was only partially considered and then neglected. The social capital in the local area, the growing size of the organisation and the related power structure embrace stakeholder engagement, influencing the orientation of the (un)dialogic dynamic. The analysis indicates that a dialogic exchange is a relative concept which depends on the interests involved and the topics discussed. It also reveals that the key factors in the democratisation of stakeholder engagement are a mutual understanding and long-term opportunities. Common sociopolitical aspects are also important, but they do not necessarily guarantee the creation of dialogism paths. The research contributes to the critical dialogic literature in revealing whether and how stakeholder engagement has been implemented in a specific setting. It also shows the limitations of voluntarist stakeholder engagement initiatives.


Introduction
Stakeholder engagement is a corporate social responsibility policy which may be used by an organisation to engage stakeholders to (un)define and (un)share solutions and outcomes (Greenwood 2007). Stakeholder engagement can be a mechanism for achieving control (Spence and Rinaldi 2014) and for accountability and consensus building (Manetti and Bellucci 2016) but also to democratise governance aspects (Brown et al. 2015;Söderbaum and Brown 2010) and to ensure co-operation and dialogue (Kuenkel et al. 2011). Often, however, engagement reflects a business-case approach (Archel et al. 2011;Georgakopoulos and Thomson 2008;Gallhofer et al. 2015) dominated by the achievement of company interests and realised through rhetorical communication (Tregigda et al. 2014).
In these cases, the interests and perspectives of the stakeholders are only marginally considered (Unerman and Bennett 2004).
In contrast with the business-case approach, critical studies have problematised the analysis of engagement, the participatory and governance processes drawing upon deliberative-agonistic democracy principles (Brown 2009;Brown and Dillard 2013a,b) and the authentic engagement process (Afreen and Kumar 2016;Bebbington et al. 2007).
They have revealed the importance of democratising the process of exchange and of reducing the power asymmetry among agents, stressing the importance of openly involving stakeholders in organisational decision making (Brown and Dillard 2015b; Bebbington 2004, 2005;Vinnari and Dillard 2016).
Drawing upon this stream of literature, this study examines the organisationstakeholders relationship by discussing whether engagement process implementation is able to promote a dialogic exchange. The research adopts a stakeholder perspective and focuses on a co-operative enterprise. A case analysis based on 16 interviews with the stakeholder representatives of a large Italian consumer-owned co-operative operating in the Italian food retail sector was carried out to investigate the topic. The literature indicates that co-operatives should be willing to implement more democratic economic exchange and socio-political practices compared to the prevailing capitalistic and neoliberal view of society, favouring pluralism and a dialogic-oriented stakeholder engagement process.
However, the democratic principles of co-operatives do not guarantee social responsibility and dialogic exchanges in and of themselves (Hernandez 2006;Mooney 2004). There may be tensions between their idealism and the business-pragmatic decisions they face (Ashforth and Reingen 2014) together with the difficulties in promoting participatory processes (Burke 2010), which overall indicate a loose coupling between the democratic principles and the actions taken.
The theoretical model integrates Bebbington et al.'s (2007) framework with the notions of power-conflict dynamics and socio-political perspectives present in the democraticagonistic literature (Brown 2009;Brown andDillard 2013a, 2015b). The study thus offers a refined framework of analysis to discuss the characterising of dialogic accounting and accountability. The empirical materials reveal a fragmented picture, where the implementation of a dialogic exchange (un)evolves depending on the interests involved and the topics discussed. While common socio-political aspects were revealed as important factors for a dialogic orientation, they were not sufficient to assure dialogism with all the stakeholders. The social capital in the local area, the growing size of the organisation and the related power structure influenced the temporalities and the dynamic of the dialogic orientation.
The present paper provides three main contributions. First, it contributes to the literature on the role of a dialogic perspective in interpreting different aspects of society and organisations, such as sustainable development (Byrch et al. 2015), public accounting reform (Harun et al. 2015), environmental planning and policy (Arunachalam et al. 2016), management control for gender equality (Wittbom 2015) and social and environmental audits (Edgley et al. 2010). Second, the focus on the stakeholder side contributes to the analysis of engagement practices by looking into the overlooked perspective of the stakeholders. The stakeholder perspective is essential for a rich and comprehensive understanding of the assemblage process which may take place among business, social, environmental and ethical as well as governance aspects (Laplume et al. 2008). Third, the indirect focus on a consumer-owned co-operative serves to raise interest in research on co-operative enterprises, which is an under-investigated setting of analysis within the social and environmental accounting and management literature (Ferguson and Larrinaga 2015;Pesci et al. 2015;O'Dwyer and Unerman 2016).
The paper is organised as follows. Section 2 introduces the theoretical framework.
Section 3 presents a literature review on co-operative enterprises. The research method is explained in Section 4. Section 5 develops and discusses the findings, and conclusions and suggestions for future research are presented in the last section.

The framework of analysis
Various theoretical frameworks and conceptual models have been proposed to discuss engagement, participatory and governance processes (Brennan et al. 2013;Contrafatto et al. 2015;Mitchell et al. 2015). The agonistic dialogic accounting literature has valorised the concepts of deliberative-agonistic democracy (Brown 2009;Brown and Dillard 2013a,b;Brown et al. 2015a) and of the authentic engagement process (Bebbington et al. 2007) to reveal the agreements/disagreements, commonalities/differences and balances among actors who interact in a given socio-political and geographical space. Its aim is to discuss the role of dialogic perspective in discussing the engagement, participatory and governance processes carried out by organisations, highlighting the importance of breaking down the powerful and hegemonic business conducts and of realising emancipatory change (Tregigda et al. 2015).
In this regard, the framework of Bebbington et al. (2007) integrated by the democraticagonistic literature (Brown 2009;Brown andDillard 2013a, 2015b) set the theoretical coordinates for studying stakeholder engagement and the related quest for participation in the present research. The framework defines seven interrelated motifs to analyse whether and how the design and implementation of social and environmental processes and tools have promoted dialogic education and created related learning mechanisms. The seven motifs, which identify critical requirements for authentic engagement, stem from the ideas of Paulo Freire, who is considered one of the most influential educationalists of the late 20th century (Contrafatto et al. 2015;Rinaldi 2013).
The first motif is the possibility for human agency. This motif underscores the notion that interactions should be constructive and open to promote different visions of society where the social needs of different actors are all taken into account. It proposes the reduction of forms of power within social relationships and, as a consequence, the creation of possibilities for sustainable living (Bebbington 2001). As indicated by Brown (2009), however, social relations are not automatically oriented to human possibilities, because power dynamics are embedded in the society and hinder the creation of emancipatory alternatives able to contrast the dominant business language. The role of power dynamics is important, because it leads to a discussion of whether marginalised groups are included in participatory processes and whether their concerns and priorities are taken into consideration in the construction of possibilities (Brown 2009).
The second motif is language and the heterogeneity of discourse, which emphasises the importance of recognising multiple voices during the design and implementation of the engagement process. This motif aims to ascertain the shift from monologic to polyvocal voices during a certain discussion (Brown and Dillard 2013a,b) in which the different actors, with their socio-political perspectives, have discussions in an open manner. This means giving more marginalised groups the opportunity to fully express their ideas. The objective is to democratise stakeholder engagement to recognise and evaluate the different values, assumptions and interests of all the different actors involved (Arunachalam et al. 2016) as well as to contrast the traditionally consensus way of promoting the engagement process carried out by organisations (Greenwood 2007).
The third motif is community and identity, which features the importance of sharing ideas and opportunities with others to create common purposes. According to this motif, the collective identity is more important than the individual identity, and the construction of a sense of community serves as the natural glue between the individual and the collective.
Establishing a sense of community based on a common identity between stakeholders may help to break down the self-celebration and self-representation typical of enterprises and in turn promote the interests of the least powerful stakeholders. Enterprises, however, have their own identity, which is often far from the needs of certain categories of stakeholders as well as from the importance of promoting social, environmental, ethical and governance practices for others (Tregidga et al. 2014). The divergent and conflicting ideological perspectives among enterprises and stakeholders and also among the stakeholders themselves indicate that the creation of a sense of community and identity is not a natural path and that, thus, dialogic forms of engagement are not necessarily constructed. In these circumstances, identifying and recognising the conflicts and struggles among actors with different backgrounds, cultures and ideological orientations may help to build up more dialogically oriented engagement (Dillard and Brown 2013a).
The presence or absence of different perspectives and languages between enterprises and stakeholders, and also among stakeholders, may be caused by the manifestation of political capital and social capital and by the creation of a "chain of equivalence". The political capital occurs when one or more groups of stakeholders chase their specific interests in contrast to the interests of the other stakeholders involved in the discussion.
When this occurs, some stakeholders obtain a specific advantage, but this also undermines the overall democratic process (Afreen and Kumar 2016;Kourula and Delalieux 2016). Social capital instead identifies a situation in which stakeholders who share the same values and visions for the future create a network based on equal and reciprocal relations to achieve common targets and realise positive externalities (Maak 2007). A "chain of equivalence" serves to align the different positions of the less powerful actors but is different from social capital. Although characterised by a different scale of values and interests, the stakeholders can create a pragmatic "chain of equivalence" to contest and change certain enterprise practices (Brown and Dillard 2013b;Levy et al. 2016). A "chain of equivalence" thus indicates the possibility "to forge links between demands that are not always obviously connected" (Brown 2009, p. 334) but which become associated temporarily to achieve a common interest.
The fourth motif is material context and power dynamics, which considers the role of power structures and dynamics within a certain context and relationship. Power is defined "in terms of the political acts of inclusion and exclusion that shape social meanings and identities" (Torfing 2005, p. 23). As indicated by Bebbington et al. (2007, p. 367), power dynamics refer to the fact that certain groups are in a better position than others to influence what are considered "legitimate", "normal" or "reasonable" ways of viewing the world and thereby impose their realities on others. As power dynamics are embedded in the society, the main challenge is how to address power relations in a way that is compatible with democratic values. Based on dialogic engagement and exchanges, the power dynamics can be reduced, and the least powerful agents can express their ideas and interact in a manner compatible with democratic values. In these cases, a space for more democratic decisions, programs and actions is created. Power dynamics, which usually characterise the relationship between organisations and stakeholders (Georgakopoulos and Thomson 2008), may thus be substituted by democratic interaction when the actors are more dialogically oriented (Brown and Dillard 2013a,b), permitting the socio-political differences to become clearer.
The promotion of a dialogic-oriented exchange may, however, cause "problems" for power elites, because it can raise stakeholder demands (Brown 2009). This is because the aim of power is to control the agenda concerning both current and potential issues (Gond et al. 2016), and thus dialogic exchanges are not an ideal option. To favour power rather than a dialogic process may be also a question of priority to make a decision more rapidly.
An example is given by Arunachalam et al. (2016). The authors show that local authorities have mixed power mechanisms and dialogic process to govern environmental planning and policy making, to rapidly come to a decision concerning the environmental sustainability of an important natural resource for the local community. Accordingly, power is a key element in terms of governance, engagement and participatory processes.
The fifth motif, institutional frameworks and democracy, suggests that the institutional framework plays a central role in driving change or, alternatively, in confirming the status quo and the dominance of some actors and ideologies over others (Harun et al. 2015). The sixth motif is epistemology, which highlights the need for a debate in which the agents involved can express their differences, conflicts and divergences without the fear of being penalised. As expressed by Brown and Dillard (2013b, p.1), "the desired outcome is not necessarily resolution of ideological differences but to imagine, develop, and support democratic processes wherein these differences can be recognized and engaged". This motif, similarly to the Language and the heterogeneity of discourse motif, underscores the importance of recognising others and of installing heterogeneous discourses (Greenwood 2007). The promotion of different perspectives and languages may expand the understanding of a certain topic, prevent particular discourses and perspectives from being dominant and, in the best cases, promote social change. However, polyvocal and agonistic debate and authentic engagement are largely hindered by the interests of enterprises (Spence and Rinaldi 2014).
The last motif is the role of experts. Who can play a role as an expert is an open question. An expert should help to open up issues for discussion, foster critical examination, guarantee access to information and improve the knowledge of the nonexperts (Brown and Dillard 2015a). How the expert exercises this role is fundamental in steering and governing the process of (un)engagement and in (un)fostering a dialogic orientation. An expert can increase (decrease) accessibility, possibilities to interact and transparent accountability for stakeholders.
Enterprises have largely operated as non-dialogic experts (i.e. they have largely operated as monologic experts who often ignore or dismiss alternative perspectives), shutting down opportunities (Brown and Dillard 2014) by implementing a top-down approach to engagement, participatory and governance processes for their specific interests (Greenwood 2007;O'Dwyer 2005). Differently, social movements (Georgallis 2016) and non-governmental organisations (Levy et al. 2016) may be experts able to inform and educate certain categories of stakeholders on how to dialogue and negotiate with companies and vice versa. Informing and educating the stakeholders by means of "bottom-up" initiatives may facilitate the identification of common orientations and targets, increasing their possibilities to influence enterprises and facilitating the production of counter accounts Vinnari and Laine 2017). Public institutions can also play a prospective role. International, national and local government agencies can stimulate a dialogic orientation and process avoiding favouring the interests of the most powerful actors when promoting debate and rule making (Brown and Dillard 2015a).
However, for what concerning corporate social responsibility issues, different and contrasting interests exist at the institutional level (Cooper and Owen 2007; Morgan 2013), which often tend to privilege the interests of the most powerful actors (i.e. the enterprises). This last motif indicates that dialogic-oriented processes can be promoted when enterprises are willing to also be non-experts, avoiding mechanisms of power and the private and public seeking for their own advantaged positions at the expense of those of the stakeholders.
The discussion presented in this section has indicated that promoting and achieving a dialogic orientation is a complex, and even radical, issue, which depends on the interplay of different elements. The analyses of whether and how a co-operative enterprise is willing and able to install dialogic engagement fit with the topic, given the democratic nature of this kind of organisation (Battaglia et al. 2016). In the next section, a literature review on co-operatives is presented to show how engagement and participation is also problematic in this kind of enterprise.

A literature analysis of the dialogic orientation of co-operatives
Co-operatives are considered to be an alternative to private and public enterprises. Cooperatives are required to operate according to the values of self-help, self-responsibility, democracy, equality and interest in the community (Birchall 2010). Attention to members' rights, employees' rights, sustainable supply chain management and local development are areas where co-operatives can demonstrate their distinctiveness and where engagement aimed at social transformation can be implemented (Battaglia et al. 2015).
Co-operatives may also have consolidated relational ties with stakeholders. Consumer, producer, worker-owned, multi-stakeholder and social co-operatives (Defourny 2010;Defourny and Nyssens 2013) are examples of the different governance models present in the socio-institutional and market spaces.
Despite the ethos of democratic principles, the literature highlights contrasting elements. Hernandez (2006) reveals how the interaction between the organisational structure, participatory culture and individual behaviour enhanced both democracy and power aspects. The author indicates that, in addition to organisational aspects, social and political relations also impacted the internal democratic governance. Such interplay created a paradoxical perspective in which both democracy and power evolved together and negatively impacted the quality of internal democratic practices. Heras-Saizarbitoria (2014) provides evidence of the internal decoupling between co-operative principles and workers' needs. The author shows that the workers' need for secure membership and guaranteed employment were considered to be the most solid political ties that bind members to the organisation. A stronger integration of the co-operative's principles on a daily basis together with a more active use of democratic mechanisms were instead less stressed, because they were considered less important. Different socio-institutional and market aspects, such as growing individualisation, precarious employment conditions, the emphasis on managerial discourse and a reduction in the social capital in the areas in which the co-operative operates, created the internal decoupling.
Ashforth and Reingen (2014) report on how the presence of two different internal groups of members, each characterised by their specific political agendas, generated a debate concerning the achievement of co-operative ideals vs. running a viable business.
The power shifted from one group to the other, and the duality of perspectives was kept continually in play through oscillating decisions and actions concerning the two different viewpoints. A democratic debate permitted the presence of the two different perspectives, which, although apparently dysfunctional at the group level, fostered functionality at the organisational level. The case of AmazonCoop Fair Trade initiatives in Latin America (Burke 2010) highlights the failure of promoting participation, democratic control and increased autonomy for the local stakeholder. While the AmazonCoop-Body Shop partnership generated material improvements for the local indigenous population, this partnership also increased vulnerability and dependency in the local indigenous population. This case is interesting because it illustrates the contradictions that emerge when a key, but not powerful, stakeholder is not democratically involved in the decisionmaking processes that concern its interests. Finally, other studies have analysed multistakeholder involvement, showing the growing importance of governance aspects (Pestoff and Hulgård 2016). For example, in the case of the co-operative Erosky, stakeholder involvement in the decision-making processes was formally guaranteed by the presence of consumer members and workers in the social council and the consumer council of the cooperative, respectively (Manetti and Toccafondi 2012). The literature indicates that dialogic engagement and participatory and governance processes are also not automatic in the case of co-operative enterprises. They must be built up and maintained both with internal and external stakeholders to favour pluralist discussions and dialogic actions. The next section describes the study's methodology and the main characteristics of the co-operative examined.

Research method
The case was designed following an exploratory qualitative analysis (Edmondson and McManus 2007) due to the relative novelty of the analysis, especially in terms of the theoretical framework. The large co-operative enterprise examined, named Alpha for anonymity purposes, is one of the co-operatives linked to the Italian brand Coop. From a governance perspective, Alpha's members possess the full ownership of the company.
The governance of Alpha followed the traditional rules of co-operatives. Through a democratic process postulated in the co-operative's charter, members vote for their representatives, who in turn elect the board of directors. The latter has control over the techno-structure, which manages the business and social responsibility aspects. Every year the members' assembly directly approves Alpha's annual budget and the annual financial statement. In 2014, the total number of members was more than 950,000. Other internal intermediate structures, such as the social council of members and the consumer council, are not present.
Over the decades, Alpha has contributed to the economic and social development of the local area. In the early 2000s, Alpha started to integrate social responsibility issues within its business model to mitigate conflicts arising from the need to operate in a competitive market. The purpose was to reinforce the connections with local communities, local actors, and members as well as secondary stakeholders through a more structured management of social responsibility and engagement. In the mid-2000s, a structured process of stakeholder engagement was implemented. From 2006 to 2011, the dialectic focused on equal opportunities, food safety, managing and valuing gender diversity, environmental protection and natural resource management.
The activities were aimed at diffusing the importance of the co-operative's principles and also at communicating the differences between the co-operative and its competitors. A series of programmes was implemented, in part thanks to the participation of the internal and external stakeholders. For example, new campaigns were set up to inform customers about product characteristics (above all regarding their safety and ethical characteristics), a photovoltaic system was installed to improve the energy management of the warehouse, and a new store layout was designed eliminating architectural barriers for disabled people (see Table 1 for the main projects implemented). For these reasons, Alpha can be considered a suitable unit of analysis for discussing the relationship between dialogic orientation and the process of engagement of different categories of stakeholders. The possibility to interview and interact with Alpha and specifically with its related stakeholders was managed through direct contact with the social responsibility manager, who had comprehensive data on the stakeholder initiatives developed and implemented over the years. This internal way of identifying the stakeholders was necessary, because the majority of the initiatives promoted toward and with the stakeholders had insufficient public information despite the presence of a sustainability report published annually. The choice of who to interview was taken exclusively by the researchers, supported by the information provided by the social responsibility manager. The researchers also directly contacted the stakeholders, with no previous introduction by the social responsibility manager.
Different criteria were used to identify the stakeholders to interview. A balanced representativeness of each category of stakeholder, following the classification of the cooperative's Charter of Values, was the first criterion. Alpha had never conducted a comprehensive map of its stakeholders; thus, the reference was the generic classification contained in the Charter of Values rather than a more precise picture of the co-operative's stakeholders. A second criterion was a pre-existent and usually long-standing relationship between Alpha and the stakeholder. Although this criterion excluded more recent relationships, which is a limitation of the research, it permitted the researchers to interview representatives with good knowledge of the social and co-operative scope of Alpha. Third, the selection of the stakeholders balanced the presence of representatives from national organisations with the presence of significant local organisations. In fact, although Alpha was one of the most prominent national market players, it has also maintained a close attachment to its historical territories. In total, 16 stakeholder representatives were interviewed (cf. Appendix A for details and acronyms used for each stakeholder group). As indicated by O'Dwyer (2005), the selection of the stakeholders is more complex than interviewing organisational constituencies, because evidence has to be collected from different groups. The relatively high number of stakeholder categories interviewed, in representation of the Charter of Values, along with the lack of an official stakeholder map as well as the discussion held with the corporate social responsibility manager led the researchers to consider the sample to be significant and representative.
The interviews were conducted between October 2013 and July 2014. Semi-structured questions were used to encourage the interviewees to take an active role in the dialogue with the interviewers but also to define the boundaries of the topics (Denzin and Lincoln 2005). The interviews ranged from 45 minutes to two hours, for a total of almost 21 hours.
The duration depended on the number of activities developed by each stakeholder with Alpha over the years. At the beginning of each interview, it was explained that the research was designed for academic purposes, thus it was not linked to the co-operative's interests. All interviewees were also informed that the interview would result in an aggregated and anonymous final scientific output. The native language of the interviewees was used for the interviews, and the research team decided not to digitally record the interviews to give the interviewees the maximum freedom of expression. While some details could not be recorded and mirroring was not possible, thus representing a limitation of the research, this method of data collection encouraged free and frank discussion on the topic (Myers 2013). Aware of the risk of losing some details, the idea was focused on "stakeholder freedom" also because, for most of the interviewees, this was the first time they had been asked for their opinions on the co-operative's actions.
The main themes discussed during the interviews were the engagement and dialogue activities carried out by Alpha over the last 5-10 years, the nature and the evolution of the relationship over time, the perceived benefits of the relationship and the problems related to the engagment and dialogue processes (see Appendix B for the structure of the questionnarie). These overlapping themes provided an in-depth perspective on the nature, characteristics and dialogic-related implications of the engagement process implemented.
The researchers took extensive notes during the interviews. On the same day as the interview, the notes were compared, and an extensive report was compiled (Myers 2013).
In addition to the interview reports, the co-operative's sustainability reports since the 2007 edition and internal documents provided by the social responsibility manager, such as the co-operative's annual sustainability plans, were analysed, thereby permitting data triangulation.
Data classification and analysis followed an iterative process. During data classification, the seven motifs were detailed to facilitate the process of analysis (Table 2) and to limit the subjectivity of the researchers. Accordingly, the materials were linked with the theoretical framework using the above-cited literature on dialogic exchanges. Then, the analysis was carried out independently by each of the authors. At the end of the process, a specific map was created to identify the significant concepts and the evidences concerning the relationship between the stakeholders, the co-operative and the theoretical framework. To establish the socio-political perspective of the stakeholders, their websites and the related available material were analysed. The socio-political perspective was interpreted considering the social, environmental, economic, political and cultural ideas.

Findings
Starting with the analysis of the possibility for human agency motif, the interviews revealed contrasting views concerning Alpha's openness and rationality in promoting a dialogic-oriented engagement. Positive claims were expressed by a non-profit organisation conditions.
An appreciation of Alpha's dialogic orientation was also revealed by a second stakeholder, S-NPO2, regarding fair trade products. Over the years, a group of consumerowned co-operatives, with the same socio-political perspective and with S-NPO2 as an advisor, built up a direct supply chain between local African producers and their Italian retail stores. A specific ethical label was designed and implemented, aimed at developing a global and socially responsible supply chain. At the time it was a niche initiative, which became increasingly adopted in the market and was partially imitated by competitors. The project linked local and international aspects. Local families, in addition to relying on millet production, received the revenues from selling green beans to the co-operative and to other local customers. This enabled the local families to go beyond a subsistence economy, with a positive impact on local socio-economic conditions. At the local level, the community also became progressively more involved in decision making concerning the project. Initially implemented for a period of almost 10 years, the project was then stopped due to supply chain problems associated with difficulty in sustaining the continuous agricultural production of the product. In 2015, the project restarted with the involvement of around 1,000 local producers organised in a local co-operative. This case revealed that, despite the partial convergence of their socio-political perspectives, Alpha and S-NPO2 Alpha understood its lack of knowledge on certain aspects and, therefore, it supported the ideas of the stakeholders as well as of the local communities.
The absence of dialogic orientation was revealed by EMP1, an internal stakeholder.
EMP1 underscored a lack of commitment toward increasing employees' skill base. At first, according to EMP1, Alpha gave more attention to promoting operational processes rather than investing in developing the skills of its workforce. For example, the investments in waste and energy management were substantial, balanced and supported by specific performance targets set by top management related to operational and normative aspects.
However, they were focused on operational procedures and internal audits, whereas related investments in training and staff skills were insufficient and not adequately planned. According to EMP1, the president and the human resource manager were unable to understand that training activities could be an important opportunity to increase middle managerial and employee skills and, at the same time, increase the commitment of the employees to the co-operative's decisions. The training programs carried out focused on transferring the operational procedures established at headquarters to individual stores (such as operative instructions regarding the separation of wastes). This was criticised by EMP1, because it had been set up without any evaluation of the needs of the employees and took a top-down rather than participative approach. Despite the numerous requests to improve employees' managerial skills, especially in the case of very large stores, Alpha's top management was not willing to interact. This lack of interaction, in the stakeholder's opinion, was not just related to training and skills but also in general regarding the need to find a better synthesis between the co-operative's principles and the market requirements (an example cited was the lack of shared managerial solutions for balancing working hours where there was a work overload at the store level). EMP1 also strongly criticised the cooperative's scarce attention to job tenure and promotion. This case illustrates that contrasting visions can exist despite the presence of a common socio-political perspective.
In this case, an internal struggle was evident and was managed following a hierarchical logic. Despite the initiatives concerning health and safety management and the diversity program slowly implemented in the stores, this internal stakeholder revealed the need to open up the internal discussion.
Other stakeholders revealed a lack of dialogic orientation. Contrasting views concerning willingness to create a heterogeneous discourse were reported by the non-profit environmental organisations. The environmental stakeholder E-NPO1 developed three projects with Alpha concerning environmental protection. In the first project, E-NPO1 gave support for identifying and designing various eco-friendly practices in the areas of energy efficiency and waste collection and differentiation. This project was externally oriented and specifically aimed at increasing the environmental literacy of the co-operative's members.
In the second project, E-NPO1 was involved in the design of new environmentally friendly products through an assessment of their environmental impacts and the consequent definition of new, specific eco-friendly characteristics. In these two examples, Alpha engaged substantially with the projects.
The third project was a collaboration in which Alpha and other consumer-owned cooperatives financially supported and sponsored a project on the massive regeneration of protected natural areas. Financial support was generated from the sale of environmentally friendly products within several stores. In this case, Alpha supported the initiative at the local level through awareness campaigns and product availability within its stores.
Despite these initiatives, E-NPO1 was sceptical about the dialogic orientation of Alpha.
This stakeholder underscored a lack of genuine openness in regard to discussing the environmental impact of the co-operative's activities. In E-NPO1's view, all consumerowned co-operatives, including Alpha, were strongly resistant to dialogue and to deconstructing their environmental behaviours as well as unwilling to question their environment-related decisions, especially when complex changes were being debated.
Some years before the interview, E-NPO1 had proposed a strong environmental repositioning requiring a significant investment in promoting environmentally friendly brands. This request involved new procedures for the sustainable management of the cooperative's supply chains to stimulate the implementation of environmental approaches in a large number of suppliers. Conflicts with Alpha emerged, because the co-operative was not willing to modify some of its suppliers' selection criteria, arguing that this could only be achieved through the simultaneous mobilisation of all consumer-owned co-operatives at the national level. In addition, E-NPO1 highlighted that dialogue and the promotion of environmental awareness with some competitors were more constructive and characterised by more openness.
For E-NPO1, the many environmental initiatives carried out over the years by Alpha, and by the co-operatives in general, were mainly driven by reputational, normative and efficiency aspects, without a genuine interest in environmental conservation. This conflictual vision was due to the fact that this environmental stakeholder and Alpha, as well as the other co-operatives, had different environmental sensitivities. E-NPO1 was a deeply green, strong, ambitious and internationally positioned stakeholder. Historically, this stakeholder had a strong interest in the co-operative's environmental decisions and performance, always requesting greater accountability. It also had much experience with business actors in raising awareness on environmental issues.
Accordingly, the relationship with this stakeholder indicates that when the environmental projects were less aligned with the co-operative's interests, the possibility of creating a heterogeneous discourse based on confrontation and participation was reduced. For example, the environment-related sponsorship initiatives did not significantly impact Alpha's internal decision making. Although oriented at promoting a more general and public awareness regarding natural local habitat conservation, this collaboration with E-NPO1 was mainly driven by a desire to improve Alpha's reputation. The respective power and independence combined with the different visions of the two agents did not favour a dialogic exchange, obfuscating the promotion of the motifs language and the heterogeneity of discourse and epistemology.
Concerning the motif of promoting a sense of community and identity, the analysis highlighted that Alpha was not sufficiently willing to create dialogic orientation with all stakeholders. With public servant stakeholders (PA1 and PA2), Alpha was able to share a common identity, because their interests tended to converge. Alpha had more than 100 retail stores in four regions. From one side, the role played by local institutions in the management of such stores was crucial (e.g. building permission and viability issues).
From the other side, the opening or enlarging of a retail store was an important moment in the life of a local community, because it was associated with new job opportunities.
Logically, the public stakeholders interviewed considered the co-operative a fundamental actor for their local communities' development. They underscored Alpha's key contributions in increasing the social and economic welfare of local citizens, offering job opportunities, developing social activities and offering good value for products. They The positive effects generated by the mutual relationship were based on a countervailing form of power (i.e. a reduction in power asymmetries) and influenced by reciprocal and common interests and by the presence of a similar socio-political perspective, which, for many years, had been a solid value between the category of public servant stakeholders and the co-operative. Despite the fact that Alpha exercised power in terms of deciding whether it was a "win-win" relationship, it also recognised the power of this specific class of stakeholders for the improvement of its business activities and territorial reputation.
However, these two stakeholders pointed out that, in recent years, the relationship had become less participatory due to a change in Alpha's position, which had become somewhat distant, less involved in the discussion concerning the development of local areas, and less willing to engage in dialogue.
As the above analysis indicates, with some stakeholders (excluding members), the development of a common identity was difficult, even impossible, due to the presence of partial and different interests. The case of the environmental stakeholder E-NPO1 was emblematic in showing cultural and practical distance. Differently, E-NPO2, a second environmental stakeholder interviewed, expressed positive argumentations toward Alpha.
E-NPO2 collaborated with several local member sections of the co-operative in the design and implementation of food security and water consumption information campaigns for the local population. This stakeholder also collaborated with Alpha's central offices, codesigning and co-implementing environmental-related projects. A prominent example was when, following the request of E-NPO2, the co-operative agreed to increase the level of environmental transparency and traceability of some of its products. In this case, the openness of Alpha was due to a convergence of interests, actions and goals with those of the stakeholder. However, E-NPO2 also recognised that the positive exchange with Alpha decreased progressively due to an increase in internal organisational complexity, which reduced the ability of the co-operative to interact actively. This case reveals a different type of interaction. While E-NPO2 was truly independent of Alpha, Alpha was dependent on the stakeholder. E-NPO2 was in fact an important national representative of consumers. Accordingly, it was important to establish a good relationship with this stakeholder to boost Alpha's reputation and give concreteness to the co-operative's principles.
The suppliers were particularly critical of the co-operative. According to the local supplier SUP2, a lack of dialogic orientation was present both at the sector and local levels. This stakeholder accused the large-scale retail market of being solely responsible for the reduction in the price of agricultural products and for the reduction in food quality.
This supplier claimed that in the retail market, the contractual dominance of large enterprises, including co-operatives, over the supply chains generated a progressive price reduction, which decreased the margins of the small suppliers. The stakeholder argued that food retail enterprises were also considered to be solely responsible for a decline in the quality of products, because they did not consider product seasonality and the fluctuation in productivity levels. This led to the introduction into the market of products from foreign countries characterised by less social and environmental regulations.
SUP2 also indicated the absence of a constructive and open relationship with Alpha, which was not interested in discussing any other issues apart from product quantity and price setting. The second obstacle was that Alpha tended to create an atomistic negotiation process with each supplier instead of designing a common and organic contractual dialectic. These relationships were dominated by power dynamics without any possibility for the suppliers to influence decision making. SUP2 indicated the absence of a common identity among suppliers as one of the main reasons for the great amount of power held by Alpha. Instead of collectively negotiating better conditions, suppliers often preferred to pursue an individual relationship with Alpha under the (mistaken) idea of being able to obtain their own specific conditions. The divergence of business interests between the co-operative and the suppliers prevented Alpha from considering the suppliers' business targets and delivery management. Dialogic orientation was absent with relationships based on the dominance of the large actor over the smaller ones.
A similar relationship was revealed by a second supplier, SUP3, who perceived a lack of willingness by Alpha to have dialogic discussions. Again, SUP3 stressed Alpha's tendency to negotiate with the local farmer confederation and to interact with individual suppliers, which, due to their small size, were unable to influence and change the negotiation process and related conditions. SUP3 reported that the lack of a common negotiation approach was detrimental to achieving better business opportunities for the suppliers. The lack of an equalitarian exchange was counterbalanced by the ability of the co-operative to pay suppliers within a short time period. This aspect clearly indicated that economic factors dominated the governance of the supply chain. Further, SUP3 criticised the lack of a common plan with the co-operative concerning the promotion of local organic products produced by small local farmers.
The engagement of suppliers was also regulated by the implementation of a food safety accreditation system. The aim of this was to evaluate quality and food security in order to officially include different suppliers within the co-operative's roster. This system was set up in collaboration with the suppliers, taking into consideration their request to be valorised as local suppliers on mass market. The idea behind the accreditation system was to offer suppliers the opportunity to sell products through Alpha's retail shops and to guarantee quality foods to the customers. This accreditation system created revenue opportunities for the suppliers, but it established a strong regime of power of Alpha over the supply chain (Spence and Rinaldi 2014).
Alpha's actions towards local stakeholders were decupled from its principles because the engagement of local and small farmers was driven by the strategic idea to establish business relations with actors in the local area and not by the idea of promoting locally developed goods and social capital. In a contradictory manner, Alpha stated that "identifying convenience and product quality, environmental protection, fair labour policies and work ethics as criteria for selecting its suppliers, the co-operative is committed to influencing the behaviour of these stakeholders as much as Concerning the role of experts taken up by Alpha, supplier SUP1, in contrast to the other two suppliers, expressed a positive indication. SUP1 explained that a long-term plan aimed at increasing its dimension and sustaining its market competitiveness during an uncertain phase of its life cycle was co-developed with the co-operative. The aim was to offer SUP1 access to the mass market. SUP1 made technological and human capital investments and successfully changed its production methods, improving the quality of its products to respond to the requests of the co-operative and of the market. The relationship maintained a dual aim for many years: sustaining SUP1's growth and offering local highquality foods in Alpha's retail stores. According to SUP1, Alpha exploited a mix of dialogic orientation and soft power, which gradually changed into a strong form of power based on price setting. This last aspect occurred due to Alpha's dimensional and geographical expansion and an increased concentration on business aspects, showing similarities with the criticism previously expressed by the other two suppliers.
A similar regressive dynamic concerning the role of Alpha as a dialogic expert was highlighted by EMP2, an employee representative. EMP2 indicated that Alpha, unlike some other co-operatives, stood by the renewal of employees' contracts as established by national law despite the economic troubles it was going through. EMP2 recognised that Alpha always considered the perspective of the employees in terms of a good salary, especially those with a lower grade of employment. However, EMP2 also revealed a gradual reduction in dialogue and face-to-face discussions. While in the past Alpha was more willing to discuss possible adjustments and improvements related to workplace conditions (such as the issue of diversity management), the nature of the relationship was becoming progressively driven by national legislative aspects (which negatively impacted the level of participative dialogue at the local level) and by the negative economic performance of the co-operative over the previous few years (which negatively impacted internal social relationships). Unlike the case of EMP1, where power played a role in resolving the divergences, in this case the weakening of the dialogic relationship followed an orderly flow. Alpha focused on economic problems that were considered more important than other ones. A similar socio-political perspective between Alpha and EMP1 was not sufficient for a dialogic exchange, because seeking a common and shared interest was subordinated to economic priorities.

Discussion and conclusions
The current study has examined the organisation-stakeholder relationship, discussing whether an engagement process was able to promote dialogic exchange. The research revealed a fragmented picture concerning the promotion of a dialogic orientation.
According to most of the stakeholders, the co-operative did not adequately promote a reciprocal understanding of the multiple interests involved but rather privileged particular discourses. Stakeholder engagement largely followed the traditional logic of stakeholder accountability (Brown and Fraser 2006), without promoting the different languages and perspectives in a dialogic manner. The co-operative thus did not set up a dialogic-oriented engagement with all stakeholders. Conversely, it often developed a consensual oriented form of engagement. Also, in those cases where the engagement of the stakeholder was democratic, it was related to the promotion of the co-operative's interests.
Concerning social issues, the stakeholders were included in the co-design and implementation of the initiatives. This occurred due to a convergence of interests and a similar socio-political perspective between the co-operative and the stakeholders involved.
In the case of environmental issues, the dialogic orientation and exchanges were flawed and almost lacking due to the presence of separate interests and divergent socio-political perspectives. Environmental initiatives, based on a consensus-oriented engagement, were aimed at increasing the co-operative's internal efficiency and normative aspects as well as reputation. They were loosely associated with dialogic orientation and loosely related to the social-political perspective of the stakeholders. The requests and the perspectives of the environmental stakeholders were mainly overlooked. In the third case, characterised by a common socio-political perspective but divergent interests, such as internal labour relations, the requests of the stakeholders were neglected because internal organisational changes were required. These three types of engagement suggest that consensusoriented dialogic orientation worked reasonably well for "win-win" relationships characterised by a convergence of interests between stakeholders and the co-operative, but otherwise was very challenging.
The analysis revealed that the weak and partial dialogic orientation was caused by the difficulties in promoting the heterogeneity of the discourse by neglecting possibilities for human agency and due to power dynamics. Although the discussion topics were aimed at proposing internal changes and reducing the business interests of Alpha, the language and the heterogeneity of discourse were absent and neglected. When the co-operative was asked to engage in organisational changes that were detrimental to its business objectivesbut concurrently able to satisfy stakeholders' intereststhe dominant profit logic did not give space for such changes (Archel et al. 2011). In those cases where the changes proposed were marginal and lateral to the business interests, the activities and projects designed and carried out with the engagement of the stakeholders (i.e. dialogic orientation) were more evident.
The case analysed did not show a commercial/moral extreme as in the case of Dey (2007), but rather it highlights a lack of perceptiveness in which the co-operative and its top managers preferred a favourable monologic perspective over a dialogic one. In fact, the discourse was dialogic only in a few circumstances despite the vast majority of stakeholders being conscious of the self-interests of the co-operative and of the necessity to find common solutions. From the analysis of the stakeholders' views, it emerges that the stakeholders were more mature and ready to be engaged in seeking out common solutions compared to Alpha, which systematically avoided conflictual situations.
While a certain level of conflict is natural between an organisation and its stakeholders (Rodrigue 2014), the analysis highlights that the creation of a common identity, as a trait important to fostering dialogic orientation, depended on the topic discussed and the willingness of the parties involved. This underscores that the notion of dialogic orientation is not a static and objective matter, highlighting that the different socio-political perspectives, power dynamics and specific interests involved in achieving dialogic orientation must be considered. As indicated by Levy et al. (2016), "parties dynamically adjust their strategies in interaction with each other and their environment. As the context of interaction changes, actors re-evaluate their opportunities and reinterpret their interests, shaping the possibilities for further moves. As a result, roles, relationships, interests and identities dynamically evolve over time". This kind of dynamism in searching for a common identity was basically neglected by Alpha despite the willingness, respect and lack of hostility of the stakeholders toward Alpha.
The analysis also revealed a decline in Alpha's stance toward dialogic orientation. The geographical proximity between the actors facilitated the development of dialogic exchange due to common cultural, political and historical aspects. The reciprocal knowledge between Alpha and some of its stakeholders reduced symbolic interaction, fostering in specific circumstances, dialogic orientation. The positive effects of the material context affected, for example, the relationships with public administrations and, for a limited period of time, also with some suppliers. Despite this, over time, the co-operative reduced the space for dialogue with its stakeholders, which some of them moved from being salient to silent (cf. Davila and Molina 2015). In the case of the suppliers, while initially the relationships were characterised by specific concessions (Skilton and Purdy 2017) based also on reciprocal needs, the power structure of the co-operative became stronger over time. With public administration stakeholders, the power dynamics were reciprocally shaped, because both the actors had the ability to exercise and influence their power over the other. The motivation to work collaboratively generated a more stable relationship, creating a more constructive exchange based on convergent interests. For the suppliers, the power dynamics became a trap, because they were no longer able (or were not willing to due to the associated economic losses) to negotiate, at least partially, the dominance mechanisms (i.e. access to mass market with revenue opportunities and diffusion of production) that characterised their relations with Alpha. These examples illustrated that in the construction of a dialogic engagement with different categories of stakeholders, it is very difficult to circumscribe power dynamics. Power dynamics can be subject to more democratic forms of accountability depending on the relevance of the topic discussed and on the nature and the role of the stakeholder involved (Boesso and Kumar 2009), but they still remain present (Brown and Dillard 2013b).
However, also for the co-operative, the strong power dynamics were not productive over time. The increase in power dynamics blocked their relationships with stakeholders. It reduced the possibility of implementing common projects, which was one of the main mechanisms that enabled the co-operative to grow in the market and to be considered an economic actor, merging economic and social improvement. The power dynamics evolved naturally and intentionally. In the first case, the power dynamics were linked to the increasingly disparate size of the co-operative and the stakeholders and to the key role played by Alpha within the local communities. The increase in size combined with a managerial lack of discernment produced a misalignment of interests between the cooperative and the stakeholders. The results thus indicated the presence of temporary seeds of dialogic orientation, stressing that the possibilities for stakeholders to open dialogues with enterprises also exist in the presence of power and knowledge barriers (Byrch et al. 2015;Levy et al. 2016).
Whether and how an organisation fosters an ongoing dialogue with a stakeholder without mobilising power dynamics is an open issue (Brown et al. 2015b). This is because an ongoing engagement with others requires a substantial amount of resources and a strong awareness and openness on the part of organisations (Parker 2014). In the case of Alpha, there was only one internal position dedicated to such an activity (the corporate social responsibility manager), without any support staff and with only partial internal sponsorship. Thus, fostering an ongoing dialogue would have been very difficult and challenging. Also, the case reveals that the development of engagement, participatory and governance dialogic processes should have a long-term orientation, whereas Alpha increasingly focused on the short and medium term rather than on the constructive management of long-term relationships (Afreen and Kumar 2016).
As indicated above, the social capital created by the historical and political affinity with similarities with what was reported by Vinnari and Laine (2013), who showed that after an initial high level of enthusiasm for environmental reporting, this turns out to be a passing fad, as was the case with Alpha's interest in dialogic exchange.
The flawed picture concerning the engagement process indicates that the adherence to business logic, characterised by the power dynamics, represented a comfortable zone for the co-operative. Conversely, opening up the discussion with its stakeholders was considered risky. In terms of Brown and Dillard's (2014) discussion, Alpha seemed to be willing to "broaden out" (i.e. consider a wide range of issues) but not really to "open up" (i.e. in being open to divergent socio-political perspectives). The declining and, more in general, the scarce interest of enterprises in the participatory rights of others generates the question of how to ensure effective engagement, participatory and governance processes. Brown (2009) highlights the possibility of creating legislative rights related to participation aspects. The results of the case analysis offer a contribution to this perspective, indicating that stakeholders do not have to be reliant merely on "voluntarist" corporate social responsibility initiatives to be engaged. Companies should be pushed to promote a more open participation and engagement through a legislative oriented approach to corporate social responsibility.
To conclude, the results of this study may be interesting for co-operative enterprise managers who are considering whether and how their organisation should promote stakeholder engagement. The results indicate that to make engagement more effective, an organisation must be strongly committed to it and have an in-depth knowledge of the stakeholders beforehand. This is because dialogic orientation is not an innate quality of the enterprise but instead represents a cultural and managerial orientation. For example, the dialogic authority of an organisation may be developed only if there is a widespread internal commitment to being democratic and to engaging the stakeholders in decision making. Concerning instead the implications for stakeholders, the creation of chains of equivalence (Brown 2009;Brown and Dillard 2013b) between the most critical stakeholders could be an effective strategy for increasing engagement with enterprises.
Political chains of equivalence may help to create mass in order to increase pressure for change and foster debate. The creation of a political chain of equivalence could promote solidarity mechanisms capable of counter-balancing the unequal power relations between the actors.
One limitation of this study is that the findings are most relevant to those consumerowned co-operatives operating with similar historical, political and organisational characteristics. A generalisation to other contexts and organisational fields must be made with great precaution. The lack of interviews with co-operative managers was another limitation. As such, interviews could have offered a complementary perspective on the topic. Due to the small number of studies on co-operative enterprises, future analyses should continue to investigate whether, how and why this form of organisation carries out engagement, participatory and governance processes. Other empirical studies based on the deliberative-agonistic literature and related frameworks could shed light on the capacity, willingness, risks and difficulties of the various organisations (co-operative, capitalistic and public organisations) to engage with stakeholders. As empirically suggested by this study, agonistic-dialogic approaches are required to take into account the more conflictual elements of pluralist relations, with specific attention to power dynamics, mutual understanding and long-term possibilities as key variables in favouring dialogic orientation.
In this regard, comparative and longitudinal analyses of different organisations could offer interesting theoretical and empirical evidence. In addition, investigating the perspectives of stakeholders concerning what the different organisations (co-operative, private and public organisations) are actually doing in terms of engagement, participatory and governance processes is a key avenue for further research. Addressing the ability of stakeholders to interact with organisations, creating for example a temporary chain of equivalence, could be an additional avenue of research, as there is no evidence in the literature on this. In fact, the assumption that stakeholders are able to engage and dialogue with organisations is basically taken for granted, and future studies are needed to explore this point. Finally, the analysis of "uninvited" forms of stakeholder engagement, rather than being reliant only on management-initiated forms of engagement, in the interest of achieving more democratic engagement/outcomes deserves future attention. Appendix B: Interview guide 1) General information about the interviewee: name, job position, employment history within her/his organisation, key characteristics of her/his organisation.
2) Identification of the nature, longevity and frequency of interaction with Alpha's representatives.
3) Specification of the involvement of the interviewee with Alpha. 4) Grade of importance -in terms of influencing -Alpha toward the interviewee's organisation.

5) Level
Alpha provides the interviewee's organisation with material or immaterial resources.
6) Level of impact of the interviewee's organisation on Alpha. 7) Evaluation of Alpha's ability to meet stakeholders' expectations. 8) Description of Alpha's impacts in a specific context with which the interviewee is familiar. 9) Level of trustworthiness toward Alpha's commitments for future interactions with the interviewee's organisation.