Introduction

Recent decades have seen governance codes mushroom as a business ethics instrument, responding to corporate scandals, transformation of ownership, and globalization. In parallel, academic interest in the topic has surged, documenting the diffusion of, compliance with, and differences between governance codes across countries (Aguilera & Cuervo-Cazurra, 2004). The origin of governance codes is generally linked with agency theory and chiefly concerned with economic objectives (Veldman & Willmott, 2016), and is often, as the nature of prescriptions, rules, and duties suggests, grounded in deontological ethics (Koehn, 1998; Rawwas et al., 2005). However, the growing burden of social and environmental objectives placed on for-profit companies (Ioannou & Serafeim, 2015) and the growing market orientation of non-profit organizations (Eikenberry & Kluver, 2004; Maier et al., 2014) make it increasingly appropriate to consider how governance codes can help organizations pursue multiple, potentially conflicting, institutional demands simultaneously.

Since corporations are marked by a separation between ownership and control (Dow, 2003), deontological rule-based ethics is seen as an apt approach to mitigate conflicts between managers and stakeholders, each pursuing their own interests (Rodriguez-Dominguez et al., 2009). Hybrid organizations, often combine multiple, potentially conflicting objectives, stemming from their aim to integrate the multiple institutional demands at their core (Battilana & Dorado, 2010; Pache & Santos, 2013). Cooperatives, for example, aim to make ownership and control fall together (Hansmann, 1999). Consequently, these organizations face unique governance challenges that are typically not anticipated in agency theory. Previous research has identified strategies and mechanisms to address these governance challenges at the organizational level, such as decoupling (Bromley & Powell, 2012), and at the intra-organizational level, such as hiring and socialization strategies (Battilana & Dorado, 2010). At the institutional level, strategies and mechanisms have so far been conceptualized only as drivers of governance challenges in hybrid organizations, not as a means to address them.

In this paper, we examine the case of governance codes devised for cooperative enterprises to understand how creating institutional prescriptions in the form of governance codes may help address governance challenges in hybrid organizations. The literature has sought to understand corporate governance codes with “large listed firms with dispersed shareholders” as both the main object of study and the normative point of departure (Veldman et al., 2023, p. 332). Complicating this literature’s focus on single logic contexts, this paper studies how institutional prescriptions such as governance codes, which seek to define good and bad behavior (Islam & Greenwood, 2021), are particularly complex to formulate for institutional environments that are characterized by a plurality of institutional logics.

By theorizing the contextual dependence of governance code’s implications for hybrid organizations, we connect business ethics to a concrete business practice resulting in both theoretical, ethical, and practical implications (Greenwood & Freeman, 2018). First, we contribute to the literature on strategies for sustaining organizational hybridity by extending consideration from organizational- to institutional-level strategies. Second, we highlight the danger associated with (mis)use of governance codes targeting hybrid organizations, in particular, the risk of institutional-level mission drift. This leaves us with the paradoxical finding that deontological ethics do not tally well with hybrid organizations such as co-operatives. Finally, we conclude with three practical implications that inform effective governance code creation for hybrid organizations.

Background Literature

Codes of Governance

A corporate governance code is a set of best practice recommendations regarding the behavior and structure of a firm’s board of directors (Aguilera & Cuervo-Cazurra, 2004), infusing deontological standards into corporate decision making (Koehn, 1998; Rawwas et al., 2005). Such standards trigger ethical actions based on duties and rules (Chakrabarty & Erin Bass, 2015). Addressing issues related to corporate decision making is increasingly seen as too complex and specific for regulation or legislation (Nordberg, 2017). At the same time, however, corporate scandals, bankruptcies, and other corporate governance failures often lead to aims such as avoiding reliance on individual decision making. This has led to rapid proliferation of these ‘binding-yet-voluntary’ codes (Nordberg, 2017, p. 79). The development of governance codes is further driven by increasingly globalized capital markets and liberalization, and transformations in ownership following increased activity by institutional investors and shareholder activism (Aguilera & Cuervo-Cazurra, 2004; Zattoni & Cuomo, 2008).

In accordance with agency theory, these drivers share the aim of increasing managers’ transparency and accountability towards shareholders (Lozano, 2000). This aim reflects either a perceived insufficiency of national regulations to protect shareholders’ interests or a desire to conform with internationally accepted best practices or taken-for-granted social expectations, and thereby to attract investors (Aguilera & Cuervo-Cazurra, 2004; Haxhi & Van Ees, 2010; Zattoni & Cuomo, 2008). For both explanations, governance codes are institutional prescriptions that constitute the observable outcomes of a negotiation process between various institutional actors (Haxhi & Aguilera, 2017; Rejchrt & Higgs, 2015). While some research has highlighted variations in these observable outcomes across countries (Heugens & Otten, 2007) and in their evolution over time (Nordberg, 2017), corporate governance codes share important similarities. While governance codes have paid attention to stakeholders and how the organization interacts with them, the literature has stressed their origins in agency theory (Klettner, 2021; Lozano, 2000; Veldman & Willmott, 2016) and its subsequent focus on addressing conflicts of interests between shareholders and managers, springing from the separation between ownership and control, by institutionalizing a set of prescriptions that guide and steer managerial decision making (Schwartz, 2005).

Next to governance codes for companies, codes targeting non-profit organizations have emerged in response to high profile scandals, including Vestia and Reinier Haga.Footnote 1 Further, governments stimulate the development and diffusion of governance code for non-profits in an attempt to modernize their governance (Walters & Tacon, 2018). While such codes are geared towards non-profit organizations, they are inspired by established corporate governance codes and are grounded in a set of “best practices” (Seidl, 2007). For example, South Africa has adopted a commercial governance code, rooted in agency theory, which was imposed on the non-profit sector (Wyngaard & Hendricks, 2010). Similarly, the Code Wijffels in the Netherlands also includes best practice and guidelines inspired by agency-based corporate governance codes (Perego & Verbeeten, 2015). Even though shareholders are absent in non-profits, principal-agent problem can manifest themselves in the relation of donors, founders and even beneficiaries with the non-profit that produces goods and services for them. The Governance Hub Code in the UK draws on aspects of the Combined Code for listed companies, including issues such as board leadership, control, performance and delegation (Dawson & Dunn, 2006). In sum, while countries opt for a single code (e.g., South Africa) or have multiple codes (e.g., Germany and Switzerland), non-profit codes typically depart from corporate governance codes which are rooted in the agency perspective (von Schnurbein & Stockli, 2013).

Institutional Prescriptions for Hybrid Organizations

Despite abundant literature on governance codes in contexts focusing on a single logic, studies evaluating this code development in institutionally more hybrid contexts, where multiple logics co-exist, are scarce. A central feature is that the logics these organizations embody are not always compatible, and may give rise to incompatible prescriptions for action (Greenwood et al., 2011). However, in the literature, the institutional context in which hybrid organizations operate has mainly been conceptualized as causing rather than solving governance challenges.

While codes of governance for conventional corporations as well as non-profit organizations offer well-documented examples of institutional prescriptions that contribute to better practices in the field (Aguilera & Cuervo-Cazurra, 2004, 2009; Cuomo et al., 2016; von Schnurbein & Stöckli, 2013), the uptake of governance codes for hybrid organizations has been slow, with evidence of especially low rates of compliance (Wigboldus & Flamman, 2018). At the same time, a well-established body of literature examines the governance of hybrid organizations, the challenges they face, and the strategies they employ to overcome these such as selective coupling (Pache & Santos, 2013), establishing spaces for negotiation (Battilana et al., 2015), hiring and socialization practices (Battilana & Dorado, 2010), and board innovation (Bruneel et al, 2020). However, these studies focus on the organizational level. Consequently, the literature lacks understanding of the role that institutional prescriptions might play in addressing rather than causing governance challenges in hybrid organizations. Furthermore, the literature on governance code development does not take into account the specificities of hybrid organizations which operate under multiple institutional logics. Hence, we ask: how can institutional prescriptions help to address organizational challenges under conditions of institutional hybridity?

We address this research question by studying the context of governance code creation for cooperative enterprises. First, we explain the background to this research context and how it relates to our theoretical interest. Next, we validate its relevance by gaining an in-depth understanding of the governance challenges of cooperative enterprises in contexts without a governance code. We then analyze the extent to which existing governance codes for co-operatives address the governance challenges we identify. In doing so, we highlight the initial, single logic-oriented scope of governance challenges considered in codes for co-operatives, and how they have evolved to address this misfit. We conclude by distilling theoretical, ethical and practical implications detailing when and how institutional prescriptions can be used to address the governance challenges of hybrid organizations.

Methods

Research Context: Governance Codes for Cooperative Enterprises

To answer our research question, we examine the creation of governance codes for cooperative enterprises. A cooperative is an enterprise owned and democratically governed by its members that pursues a purpose defined by its member base to address their needs, rather than maximization of shareholder value (Novkovic, 2008). Cooperative members may wear multiple hats, for example members simultaneously being worker and manager in worker co-operatives, since they are set up with an aim of making ownership and control fall together (Levi & Davis, 2008). Traditionally, four types of cooperatives are distinguished, based on the type of member-owner: worker cooperatives, consumer cooperatives, producer cooperatives, and multi-stakeholder cooperatives (Eum et al., 2020). Cooperatives are strongly represented in various sectors of the economy, including agriculture (Ménard, 2018), mutual insurance (Hansmann, 2000), and renewable energy production and distribution (Bauwens, 2016).

Building on previous research that conceptualizes cooperatives as hybrid organizations (Battilana, 2018), we identify them as integrating three institutional logics and their related demands (see Fig. 1). First, the economic logic of cooperatives relates to the fact that they produce services or goods and operate in markets, and thus need to be financially viable. Second, their democratic logic links with their aim to be self-governed by their members on a ‘one member, one vote’ principle, rather than the conventional link between capital (shareholders) and control. Third, cooperatives are set up to fulfill certain needs or goals for their members, an aspect we refer to as their purpose orientation. In combining their mission-driven nature and their democratic decision making while operating in a competitive market (Mitzinneck & Besharov, 2019; Van Oorschot et al., 2013), cooperatives face specific governance challenges (Ménard, 2004; Nilsson, 2001).

Fig. 1
figure 1

Cooperatives and their governance challenges: Tensions within and across logics

Full size image

At the institutional level, support practices have been developed to help cooperatives to manage these challenges. Examples include monitoring and certification systems through national initiatives, such as the National Cooperative Council (NCR) in the Netherlands and the National Council for the Cooperation (NRC) in Belgium (Cracogna et al., 2013), and increasingly through dedicated codes of governance. Cooperative federations, which are often national umbrella organizations, have been developing governance codes for cooperatives. The first was issued in 1996 in the UK, followed by the Netherlands in 2005 (Co-operatives UK, 2013b; NCR, 2015). More recently, the UK has integrated three earlier governance codes for agricultural cooperatives (Co-operatives UK, 2013a), worker cooperatives (Co-operatives UK, 2012), and consumer cooperatives (Co-operatives UK, 2013b) into a single governance code comprising the entire cooperative sector (Co-operatives UK, 2019). The Netherlands recently published a fourth revision of a sector-wide governance code (NCR, 2019), and other countries such as Sweden have also established a sector-wide governance code (Svensk Kooperation, 2019).

Research Design and Sampling

Given the exploratory and processual character of this study, we adopted a case study-based research design with a grounded theory approach (Charmaz, 2005; Glaser & Strauss, 2006/1967). We addressed our research question through multi-level data collection, iterating between data, preliminary analyses, and theory. Since the literature focuses on contexts with a single institutional logic, we contend that our question of what governance codes should look like, can only be addressed in an empirical context characterized by a plural institutional environment.

In order to explore the functionality of governance codes in a plural institutional environment, and crucial for examining complex unfolding phenomena (Gehman et al., 2018), our process model requires cases that show multiple phases of the same processes (Langley & Abdallah, 2011). Tying together the various national contexts, we thus sample so-called “strong” cooperatives that all adhere to the principles of the International Cooperative Association (ICA) (Fici, 2013), but are different with respect to the presence or absence of a governance code. This adherence to ICA principles can be facilitated either through the locally adapted legal frameworks or through the use of substitutes like national associations which seek to implement these principles (Fici, 2013). In doing so, we intend to increase the internal validity of the processes we theorize. Indeed, ICA is a vibrant and influential global movement (Mushonga et al., 2019) organized around multiple sectoral and geographical apex organizations, and actively promotes the distinctive cooperative identity, values and principles (ICA, 1995). While for each of these national contexts also “weak” cooperatives exist that are cooperative only in name or legal status, our sampling strategy assures we are looking at cooperatives that, despite different national and jurisdictional contexts, share the same ICA-inspired identity. This ensures that no organizations in our sample have a superficial implementation of the cooperative identity and principles and thus an absence of its specific governance challenges (Fici, 2013). Next to sampling on “strong” ICA-adherent cooperative movements, we sought to select cases that enabled us to study different phases of the same process. In particular, we selected a national context in which the creation of a governance code was being debated (Belgium). This enabled us to collect data at the organizational-level to understand the sorts of challenges they were seeking to address through such governance codes. Next to this, we studied two national settings (the Netherlands and the United Kingdom) that had extensive experience with governance code creation, which allowed us to look retrospectively at how codes had sought to address these challenges.

The empirical context of Belgium allowed us to identify the governance challenges that stakeholders wished to see remedied by a code, a crucial first phase in the process. In a context with a governance code already in place, ex-post rationalization by actors would have made it hard to disentangle the drivers of code creation (Aguilera & Cuervo-Cazurra, 2004). Within Belgium, we used a theoretical sampling strategy based on a stratified purposive sampling approach (Draucker et al., 2007; Patton, 2002), maximizing variation in both context and actors (Miles & Huberman, 1994), as long as our sample cooperatives matched our overarching sampling strategy of picking “strong” cooperatives. Since cooperative codes of governance unfold in all strata of the cooperative sector in Belgium, we sought a sample that would capture maximum variation in terms of different types of cooperatives. To do this, we adopted four main sub-criteria for which theory shows variation in cooperative governance issues: type of cooperative (Eum et al., 2020), maturity (Cook, 2018), sector of activity (Bijman et al., 2014), and size (Chaddad & Iliopoulos, 2013). We also maximized variation in the types of actor researched (Miles & Huberman, 1994), and engaged with at least one member and one manager of each cooperative in our sample. The members’ involvement in the cooperatives ranged from regular members to volunteers, ambassadors, and board members. This enabled us to embrace multiple viewpoints on governance within the same organization to triangulate our understanding (Miles & Huberman, 1994). Table 1 provides an overview of our cases and their attributes at the organizational level.

Table 1 Attributions of organizational interviews
Full size table

To gain retrospective insight into how codes are created, our second context included multiple institutional-level actors in two countries, the UK and the Netherlands, that had already implemented cooperative codes of governance for their “strong” ICA-adherent cooperatives, in 1996 and 2005 respectively, thereby offering an appropriate empirical context to map subsequent phases in our process study. While other national movements, like Spain, Italy or Poland also know “strong” ICA-adherent cooperative movements, our data collection took place in the UK and the Netherlands, because of their extensive experience in developing and revising governance codes. This sample allowed us to carefully map the code creation process in cooperative sectors that have already undergone multiple alterations and changes. This provided us with an opportunity to uncover apparent misfits between governance codes and governance challenges. Table 2 provides an overview of our cases and their attributes at the institutional level.

Table 2 Attributions of institutional-level interviews
Full size table

Data Collection

We constructed a data set from multiple data sources relating to our cases, including both primary and secondary data (Yin, 2003). Our primary data consisted of interviews and feedback sessions with key figures in the cooperative sector while our secondary data consisted mainly of archival material, publicly available codes, and existing questionnaire data. In the first round of data collection, from the end of 2019 until early 2020, we collected data at the organizational and institutional levels in parallel, resulting in constant iterations between theory development and raw data. During this period, 42 interviews were conducted with individual cooperatives and key institutional figures. The interviews lasted between thirty minutes and three hours, with an average of one hour. Topics covered at the organizational level included allocations of rights and responsibilities, challenges of decision-making bodies, and the kinds of issues they wanted to see addressed in a governance code. Institutional-level interviews revolved around the initial motivations for code creation, how it was received by the sector, and ways in which such feedback led to revisions of the code. In addition to primary data extracted from the interviews, 28 documents were reviewed as secondary data. These included various governance codes from the UK and the Netherlands, and meeting reports, draft versions, and presentations to develop a better understanding of the code creation process. Our secondary data set amounted to a total of 765 pages.

Following our preliminary analyses of how codes of governance were created in our empirical context and how they tackled typical governance challenges, in 2021 we conducted a second round of semi-structured interviews with some of the same institutional-level respondents to fill small gaps in our framework, collect feedback, and nuance our analysis. We also validated our emerging theoretical insights by checking with members in a feedback session with key figures in Belgium’s cooperative sector (Lincoln & Guba, 1985).

Data Analysis

Stage 1: What Governance Challenges Do Cooperative Enterprises Face?

Drawing on data from interviews with actors in the cooperative sector in Belgium, this stage of analysis had two goals: to validate the relevance of the theoretical understanding of cooperatives as hybrid organizations that we had derived from existing literature, and to gain an in-depth understanding of how simultaneous pursuit of competing institutional demands creates governance challenges for cooperative enterprises. To do so, we iterated extensively between theoretical concepts in existing literature and our empirical data.

Stage 2: Understanding the Structure and Content of Cooperative Governance Codes

The purpose of the second stage of the analysis was to gain a better understanding of governance codes for cooperatives in the UK and the Netherlands.Footnote 2 In our analysis of publicly available governance codes, we focused on similarities and differences between the process through which governance codes have been created in these countries over time. Meeting reports of the commission charged with developing the governance code and preliminary versions of the codes enabled us to grasp the rationales for these changes. Next, we analyzed the transcripts of interviews with three field experts in the UK and the Netherlands (I1, I2, and I3) to validate and deepen our understanding of contextual factors and motivations affecting the governance code creation process.

Stage 3: How Can Governance Codes Help to Address Governance Challenges of Cooperative Enterprises?

In the final stage of our analysis, we sought to integrate the findings from the first two stages by adopting a theory-building approach (Glaser & Strauss, 2006/1967; Langley, 1999; Orton, 1997). The interview transcriptions were approached inductively in a first coding round in vivo (Saldaña, 2009) using NVivo software. The data were gradually grouped into coding categories reappearing in various interviews, and these in turn were grouped together into the various dimensions of the code creation process. This stage sought particularly to address whether, when and why cooperative governance codes address specific governance challenges. We constructed a theoretical rationale to explain variation in governance codes based on their (mis)fit with the governance challenges we had identified at the organizational level. To validate our emergent theoretical insights, we performed a final round of follow-up interviews (I4 and I5). During this phase, additional data were also collected on the uptake of newly implemented governance codes. This analysis led to conceptualization of two distinct approaches to code creation, termed ‘adaptation’ and ‘co-construction, discussion of which forms the basis of the second and third parts of our findings.

Findings

Governance Challenges in Cooperative Enterprises

In analyzing both actual and perceived governance challenges of cooperative enterprises based on our interviews with members and managers, we distinguish two key aspects. First, cooperative enterprises are subject to tensions within logics, or challenges associated with implementing practices aimed at achieving goals associated with a particular institutional logic (see Table 3). Second, additional tensions arise across logics, owing to simultaneous pursuit of multiple institutional logics, and contradictions in implementing the practices necessary to achieve them (see Table 4).

Table 3 Overview of institutional logics and their challenges
Full size table
Table 4 Overview of tensions across institutional logics
Full size table

Tensions Within Logics: Maintaining a Shared Purpose, Achieving Financial Viability, and Achieving Democratic Control

Cooperative enterprises develop around a shared purpose, categorized as either a mutual or a general interest orientation according to the extent to which members and beneficiaries overlap (Bauwens et al., 2019). With a mutual interest orientation, members often join forces in a cooperative to reach a certain scale, thus ameliorating their market position in the face of market failure. However, this orientation may also benefit members in non-financial ways, such as by creating autonomy for workers in a worker cooperative, as illustrated by the following quote:

That’s just the advantage of something like [a cooperative], otherwise you have to motivate people by paying them more. Everyone gets paid well – they should – but that’s not the motivation. That’s a whole different approach to economics and work. (M1)

In contrast, a general interest orientation refers to members’ desire to create value for non-members and/or to produce change at the societal level:

[The members] were favourable to the idea that there would be employment, for example, for people who had just left prison. You are not buying chips; you are buying a social outcome, you are buying a social impact. Some of the cooperators bought shares to promote veggie and vegan food to non-vegetarians. (W4)

Rather than choosing one or the other, the flexibility of the cooperative form is often leveraged to find a balance on the continuum between mutual and general interest. Nevertheless, a shared understanding of this sense of purpose must be created and sustained among members to avoid value clashes and enable effective organizational action (Dufays, 2019).

The economic logic of cooperatives implies that in addition to their purpose, cooperatives need working business models to achieve financial viability. This viability may be affected by both internal and external factors. Internal factors relate to attracting the resources and skills needed to manage and sustain daily operations. This respondent illustrated that this need for resources and skills heavily depends on the sector, scale, and diversity of operational activities. In contrast, external factors capture the effects of industry dynamics on cooperatives’ internal operations, such as competition, consolidation, and digitization. Integrating external and internal elements is essential for managing tensions relating to the economic logic, and resembles the traditional corporate governance challenges of investor-owned firms. Maintaining financial viability, economic competitiveness, and a professional board composition are considered key issues in this respect.

Cooperatives are democratically controlled by their members and can embed member participation through either a direct or indirect model. The direct model refers to settings in which members relate to each other on a basis of trust and reciprocity: ‘This is where participation and control often coincide. Because that is actually a very small entity that interacts with each other in a very interactive way’ (W4). This highly interactive setting leads to low levels of asymmetric information and high degrees of member participation. The often ad hoc and informal decision-making process is non-hierarchical, involving all members either at the same time or through a rotation of roles. Key barriers to democratic governance in direct models relate to informal and group dynamics, as well as the selection and socialization of new members. In contrast, the indirect model refers to larger and often more heterogeneous member bases, where the democratic process is more formalized and structured into (representative) decision-making bodies. Since sustaining informed and active member participation in these settings is complex, often only a subsection of the member base is actively involved in the cooperative enterprise’s governance. In indirect models, structuring the process of democratic control by a larger member base is the key challenge, and includes issues such as sustaining member participation, and selection procedures for decision-making bodies.

Tensions Across Logics: Business–Purpose, Individual–Collective, Expertise–Representation

The degree to which challenges manifest themselves across the economic logic and the purpose orientation of cooperative enterprises varies widely based on the interest orientation pursued. First, cooperatives characterized by a strong mutual interest orientation face the fewest problems along this axis, owing to greater alignment of the goals pursued under the two institutional logics. Second, cooperatives with a strong general interest orientation or weak transaction relationships with members face stronger tensions along this axis. In particular, finding a balance between a general interest orientation and financial viability may prove to be profoundly challenging. Insufficient attention to the economic logic may cause these cooperatives to struggle to achieve financial viability. The opposite situation, in which the cooperative’s economic logic takes precedence over its purpose, is denoted by mission drift, a well-documented phenomenon for hybrid organizations in general (Ebrahim et al., 2014; Grimes et al., 2019). This challenge is particularly salient for more mature and larger organizations and those with lower levels of member participation:

If that organization has been in existence for a while, I see that all kinds of dynamics get going within that organization that actually deviate from the basic mission, vision, and objectives and actually aim to keep that structure alive. (C10)

At the intersection of the democratic logic and its purpose orientation, cooperatives must balance individuals’ and subgroups’ incentives for participation with the organizational-level purpose, thus revolving around a number of social dilemmas. Manifestations of this challenge include informal hierarchies relating to the founders’ status, experience, and personality in small cooperatives:

Those ancients have always been a difficult case. I think those five ancients always opened things up, but it was not always perceived that way by others. The informal authority that the five ancients had was still considerable. Also, because those five were the only shareholders for a very long time, we can’t get that out of them. (W5)

Among larger cooperatives, this challenge relates more to incentivizing and sustaining participation among a base of members who increasingly act as regular consumers:

That [turnout at the AGM] is very low. Everyone [120,000 members] is invited, but only about ten people come. One of the problems is that most people are members, but they don’t identify with it … In the end, the financial participation is only 2.5 euros, it’s not the world. (C4)

Actions prescribed by the economic and democratic logics give rise to conflicting tensions in various situations:

For the board of directors, it is a bit of a search. On the one hand, there must be enough skills, but on the other, the feeling with the [member] base must also remain. I don’t know the best way to express that. (M4)

As this quote illustrates, selecting decision makers requires consideration of issues relating to the economic logic, which stresses the importance of expertise, efficiency, and ability to closely monitor different levels of management. However, the democratic logic stresses aspirations for self-governance and avoidance of hierarchies through rotations or representation. Neither issue can be entirely neglected in a cooperative enterprise. According to our respondents, this requires considerable effort to balance sufficient contact with members on the one hand, and sufficient managerial skills on the other, to reconcile these tensions across logics.

For cooperatives that can navigate these tensions, this triple-logic constellation may be a virtuous circle. Members raise questions and generate ideas in an informed manner. When these suggestions are found to be in line with the organization’s shared purpose, the enterprise starts to look for a financially viable way to provide for these member needs, which in turn strengthens member participation. However, the outlined tensions within and across logics (see Fig. 1) imply that cooperative governance is about continuously balancing conflicts and deviations that may upset this triple-logic constellation.

Creation of Cooperative Governance Codes: Two Mechanisms

Turning to the actual creation of governance codes, we uncover two processes which we label ‘adaptation’ and ‘co-construction’ (see Table 5). In both our empirical contexts, the initial governance code creation followed an adaptation mechanism. However, as these codes were marked by low uptake and superficial compliance, they had been revised numerous times. In the Netherlands, the last revision led to a drastic revision of its governance code. In what follows, we theorize how these differences led to stark variations in content and compliance, and ultimately in the effectiveness of these two governance code creation mechanisms in addressing the challenges associated with hybrid organizations.

Table 5 Overview of the two mechanisms for code creation
Full size table

Adaptation

Our analysis suggests that code creation through adaptation has been inspired by developments in corporate governance codes for investor-owned companies, and has been introduced as a result of similar triggers, such as corporate scandals. These pressures are nicely illustrated by the process through which the UK’s agricultural cooperative code of governance came about in 2013 (Co-operatives UK, 2013a). First, a parliamentary commission assessing the failure of the Dairy Farmers of Britain cooperative recommended the development of a code of governance to promote good governance practices in the sector. This proposal was accepted by the UK government, which committed to working with Co-operatives UK, the national umbrella organization, and the Scottish Agricultural Organization Society (SAOS) to achieve this. In the Netherlands, similarly, the trigger for cooperative governance code creation was not demand from the sector itself, but rather development of a corporate governance code for investor-owned companies:

At the time, the Dutch code Tabaksblat, as it was called, was actually the first corporate governance code developed with the rules of how a corporation worked, really the current rules actually. And because we felt that the cooperative was different from the traditional company in a number of ways, it was felt that we needed to create our own cooperative code. (I3)

In addition to the reason for creation, our analysis indicates that inputs and processes in creating these codes were also largely external to the cooperative sector. This is illustrated clearly by the creation of the 2013 governance code for consumer cooperatives in the UK (Co-operatives UK, 2013a, 2013b), for which an external consultant, with experience in governance codes for publicly listed companies, was asked to ‘draft the skeleton of a code’. The cooperatives themselves ‘were adding those cooperative elements into it’ (I2) since the consultant had very limited knowledge of the cooperative model’s specificities. During the process of code development, members’ participation was intentionally limited ‘because if you consult too much with the people that you are going to expect to follow it, then they are going to water down what the codes says’ (I2). A similar development process in the Netherlands turned the code into a traditional corporate governance code, with several elements deleted or added to fit the specificities of a cooperative enterprise:

But the process was, we have a code, and the first code there was was very much based on the Tabaksblat code in 2005 … We will look line per line: does that apply to us or do we think it is different? (I3)

Cooperative governance codes following this creation mechanism are ‘not necessarily massively different’ (I5) from governance codes for publicly listed companies which operate under a dominant economic logic. Therefore, they often fail to address issues relating to other institutional logics or challenges relating to tensions across them. The envisaged contribution of this type of code, in line with the economic logic, is to ensure the cooperative’s accountability to shareholders by ensuring that the board of directors is sufficiently skilled to fulfill its role of monitoring and controlling management. In doing so, the main problems recognized are lack of board member rotation and lack of professionalization in terms of skill sets.

Based on the preceding analysis of governance challenges in cooperative enterprises, we would expect this inherent prioritization of one logic over others to be strongly contested. Indeed, we find that organizations that did not share this prioritization followed one of two courses of action. First, some enterprises did perceive these governance codes as undermining their hybrid identity, and opposed them:

In the code this time we are ensuring that directors are elected on some kind of competency-based formula. Some coops absolutely see this as undemocratic even though it is in the code. There is one big coop, not the biggest but a big one, that is really opposed to anything like that. (I5)

Second, and most importantly, a large proportion of the cooperative sector decided not to adopt these codes because they ‘lacked specific cooperative elements’ (I3) and did not resonate with the governance challenges faced by cooperatives. This was especially true for cooperatives in which the economic logic was less central: ‘We also had people from energy cooperatives and care cooperatives, from smaller cooperatives that did not recognize themselves sufficiently in the old code’ (I3). Moreover, the governance code had ambiguous implications even for cooperatives that did adopt it, as compliance was superficial. First, respondents highlighted a ‘box ticking’ (I3) mode of compliance that had little effect on these organizations’ concrete governance practices:

There were some where it seemed like more of a race to the top, who could comply with the most, which obviously wasn’t what it was all about. And there were others whose governance perhaps wasn’t quite as they were saying it was when they said they were complying with things. (I5)

Even worse, as the code seemed to define best practices without consulting the sector, it seemed to be decreasing scope for diversity in the sector’s practices. This was because individual cooperatives now wanted to ensure that their practice was considered ‘best practice’, and thus sought to delegitimize different practices by other organizations:

Another thing we have seen which hasn’t been always positive is more challenge between these coops. Where one coop sees the practice of another one which they do not think meets the standards set out in the code, then that is something they come to us about and say ‘they are doing this, this and this and we don’t think they should be doing it’. That’s good but not always positive, because often it is just that they are doing something different from what their coop is doing, so they think that must be wrong because their coop is obviously always right. (I5)

The last consequence of this prioritization relates to the risk that governance codes seek to push cooperative enterprises towards more standard corporate governance practices. The scale at which they do so threatens the very identity of the sector and risks collective mission drift. Indeed, organizations clearly identified ‘what you [Co-operatives UK] are doing’ (I5) as threatening their hybrid identity by ‘actually undermining democracy’ (I5). As ‘the new practices look more like what we would call a PLC [public limited company] model in the UK’ (I5), some cooperatives just did not ‘want to see those kinds of governance practices creeping in’ (I5).

However, in both countries, low uptake and superficial compliance led to a series of revisions and updates (see Fig. 2). Throughout these revisions, considerable emphasis was given to reducing rigidity, complexity, and overly detailed accounts of specific governance situations. In the Netherlands, this process went beyond incremental adaptation, after a formal monitoring study found that 91% of the members of the cooperative federation were not reporting on their compliance with the governance code (Wigboldus & Flamman, 2018). The study concluded that ‘especially the cooperatives of limited size seemed to have difficulties in complying with the 72 rules of the Code. As a consequence, the NCR Code of 2015 seems to insufficiently connect to the diversity of NCR’s members’ (Wigboldus & Flamman, 2018). Hence, the 2019 revision of the governance code started out with the ambition of creating a new type of governance code that would address these issues.

Fig. 2
figure 2

Evolutionary overview of cooperative governance code creation in the Netherlands and UK

Full size image

Co-construction

The co-construction mechanism for governance codes was a response to low adoption and superficial compliance of the early codes. It was pioneered in the Netherlands in 2019 (NCR, 2019) through a process that aimed to stop producing ‘an update of the old code’ (I1), and instead ‘rewrite it from A to Z’ (I1) because ‘it has to be different’ (I1). In the process, the code was shortened by 75%. Moreover, in contrast to governance codes based on adaptation, this mechanism did not take corporate governance codes as a starting point:

We actually set out with a couple of starting points. The first principle was: everything that’s already in the law, we’re not going to repeat. The second is we are only going to write down what makes the cooperative unique from other businesses. (I3)

Similar to code creation under adaptation, the process involved some governance experts, for example academic scholars. However, to identify their unique challenges, the NCR brought together a sample of its members, varying in size, sector, and activities, to engage in a deliberative process to co-construct several principles that would fit the entire cooperative sector. Greater attention was given to ensuring proper representation of diversity in the field: ‘Until 2015, we always had people from larger cooperatives. Now we also had people from energy cooperatives, from healthcare cooperatives, from smaller clubs that didn’t recognize themselves enough in the old code’ (I3).

Starting points for discussion were provided by existing principles, definitions of cooperatives’ identity, and other institutional prescriptions that had been devised specifically for cooperative enterprises:

So among other things we looked into the ICA principles, looked at the principles of Raiffeisen; what makes a cooperative unique and to what extent this can be expressed in principles. (I3)

The explicit focus on distinctive aspects of cooperatives as an organizational form led to a stronger link with their hybrid identity in the governance code. Indeed, the basis for the code’s applicability resonates with their triple-logic constellation: ‘In all instances it concerns an association [or democratic logic] with a company [an economic logic] and in all instances there is a shared vision [a purpose orientation] as an underlying foundation’ (NCR, 2019). This shift to focusing on distinctive aspects of cooperatives had profound implications for the goals the code was expected to achieve, namely it ‘aims to stimulate conversation between cooperatives amongst each other and in cooperatives internally’ (NCR, 2019). Rather than defining best practices, monitoring them, and hoping that this would strengthen cooperatives’ accountability to members, the focus of these governance codes was on supporting maintenance of their hybrid identity. This shift in focus led to very different questions being asked:

How do you activate members? How do you make sure member participation is being shaped? How do you make sure that the opinion of members is taken into account in decision making? These things are unique for the cooperative. (I3)

In this respect, the code was considered ‘successful if it had contributed to awareness of the ideas behind the principles and provisions and if it led to vibrant dialogue within and outside of the cooperative’ (NCR, 2019).

As this goal is obviously more difficult to monitor, these governance codes are themselves principle-based rather than rule-based, and allow for diverse interpretations and implementations. Their key objective is to get members to discuss and translate these principles into their own organizational goals, activities, and governance processes, which makes their adoption a more active process, and monitoring their compliance both harder and less central. In this sense, the co-constructed governance codes help to deepen the cooperative identity through discussions among members, a process referred to as ‘cooperative dialogue’ (I3).

Interestingly, this cooperative dialog did not consider governance challenges only within the economic logic, nor did it inherently prioritize one logic over others. Consequently, tensions across the economic logic and the cooperatives’ purpose orientation were explicitly discussed:

You see that for a number of cooperatives the economic was very dominant, while for others it was more the social and societal objectives. What makes a cooperative unique then? Is it that everyone includes both elements, or can you also choose between them? These are things that were very different between cooperatives. (I3)

Tensions arising between cooperatives’ economic and democratic logics were also discussed:

So a cooperative is an association with a firm. What you often see is that the firm develops strongly and that the member side is actually underdeveloped, and that members act much more as customers than as co-owners, so you have to reactivate that, and we think that the code also helps to reactivate that sense of ownership. (I3)

Lastly, trade-offs were considered between individual and collective interests that might arise at the intersection of the purpose orientation and the democratic logic:

What does it mean for my cooperative if they say that solidarity is one of the principles? How solidary are we and for whom? (I3)

If we have 1,000 members and we are making policy for the coming years, are we then trying to make it for these 1,000 members, or for the top 90%, or for the top 80%? … And if we choose the 80%, what are we going to do with the other 20%? Are we going to try to let them go or are we going to do the opposite? (I3)

Discussing these questions aimed to create a sense of shared understanding of priorities, trade-offs, and avenues for integrating conflicting objectives, at the sectoral level in the code creation process, and at the organizational level as they sought to actively implement these principles. This was likely to reduce contestation while increasing scope for diverse and innovative practices. More concrete guidance in the form of practices, models, and structures remained available to organizations, albeit in a more suggestive manner. This was achieved by complementing this short, principle-based code with a dynamic, digital database of best practices in the sector. Recognizing that these practices did not necessarily fit everyone, they were ‘meant as an inspiration, as a lever, as good practices, to support the cooperative’ (NCR, 2019), rather than being prescriptive.

Discussion

In this paper we ask how institutional prescriptions devised under institutional hybridity might help to address governance challenges for organizations operating within such contexts. To answer this question, we investigate governance codes as a concrete example of institutional prescriptions. We examine how they can address the governance challenges of hybrid organizations do not fit the classic governance challenges identified by agency theory, in two steps. First, we theoretically conceptualize cooperatives’ hybridity by reviewing existing literature, and empirically validating and deepening our understanding of the governance challenges they face. Second, we examine existing governance codes to understand their (mis)fit with the governance challenges identified. In this section, we outline the theoretical, ethical and practical implications resulting from our analysis.

Theoretical Implications: Sustaining Hybridity from the Outside

First, we extend the literature on hybrid organizations’ governance challenges. The literature focuses on strategies that address governance challenges at the organizational and intra-organizational levels. Complementary to these agentic solutions, we highlight how institutional arrangements, by devising institutional prescriptions such as governance codes, can address the tensions pertaining to conflicting logics at the organizational level.

Our results show that governance codes may facilitate negotiation of a shared understanding of values and principles within the hybrid organization. The governance code may be used as a socialization tool to reduce organizational members’ divergent understandings of the institutional logics at stake, by clarifying expectations stemming from individual logics and their combination (Meyer & Höllerer, 2016). In this regard, it helps to create a negotiated prioritization for the interrelation between logics. This is especially relevant when conflicting institutional logics are found to be central to the organization’s functioning, a situation referred to as contested hybridity (Besharov & Smith, 2014). In such cases, governance failures may be caused by conflict arising from differences in their prioritization among members of the organization.

The second potential role of governance codes in supporting organizational hybridity is relevant when conflicting objectives’ logics differ in their centrality to the organization’s functioning. In such cases, organizations face the risk of estrangement from the less central logic and thus from their hybrid character (Besharov & Smith, 2014). To address this issue, governance codes offer concrete practices, models, and structures that allow organizations in a context of estranging hybridity to re-embed the institutional logic at risk. The governance code may be a useful tool how to implement the negotiated prioritization of logics. Rather than distinct logics imposing distinct demands on organizations, the code helps them from drifting away from one of the logics making up their hybrid identity.

Ethical Implications: The Danger of Institutional Prescriptions for Hybrid Organizations

Our analysis suggests that institutional-level strategies to sustain organizational hybridity act as a double-edged sword. A successful governance code implementation has the potential to strengthen the sector’s hybrid identity and to help it mitigate the challenges of contested and estranging hybridity. How to simultaneously integrate social and economic performance has been extensively studied as an ethical challenge at the organizational level (e.g., Eikenberry & Kluver, 2004; Ramus & Vaccaro, 2017; Staessens et al., 2019). This paper theorizes how institutional-level prescriptions, such as governance codes, could be seen as a complementary way to resolve this challenge and combat mission drift in these settings. At the same time, institutional-level prescriptions can also be defined in a way that misfits the hybrid identity. If this is the case, its impact might be limited by low uptake or superficial implementation.

However, when governance codes that present a misfit are diffused within one sector, these institutional prescriptions bear not only the risk of causing mission drift on the organizational level but could potentially cause mission drift for an entire sector, which we label institutional-level mission drift. In a context of contested hybridity, a misfit could lead governance codes to impose a ‘best’ practice in the sector, implying a prioritization of institutional logics that might not be shared by the organizations operating within it. In line with our analysis, the increasing opposition and conflict both within organizations and in the sector more broadly is likely to diminish the scope for diversity within the sector, to hinder diffusion of innovative practices, and ultimately to risk undermining the sector’s resilience. With regard to estranging hybridity, imposing a prioritization among institutional logics and the practices associated with them may be even more disastrous. In the cooperative sector of our study, prioritizing the economic logic over concerns for democracy and purpose has led to an institutional-level mission drift, as the practices put forth by the governance code are seen as incompatible with other institutional logics relating to their hybrid identity.

Paradoxically, while deontological ethics have often served as the basis for code implementation (Koehn, 1998; Rawwas et al., 2005), and have especially guided the cooperative movement through a set of internationally accepted standards and values (Spear, 2004), our findings suggest that a deontological take on cooperative governance could even actively drive the sector as a whole towards mission drift. While governance codes are meant to avoid potential conflicts between all sorts of stakeholders on the one hand (of which specifically shareholders), and managers on the other hand (Rodriguez-Dominguez et al., 2009), these potential conflicts of interests are typically informed by agency theory. However, cooperatives aim to make ownership and control fall together which contradicts the core tenet of the agency perspective (Dow, 2003; Hansmann, 1999). Moreover, our data illustrates how these organizations are often set-up by their members to pursue non-economic interests or even have the explicit purpose of benefitting others. However, this is another source of tensions in cooperative governance that is not anticipated by agency theory. Our findings, then, seem to reveal an inconvenient but surprising paradox with respect to the governance of so-called hybrid organizations such as cooperatives: deontological ethics seem to be in opposition to the cooperative reality, which longs for more pragmatism when it comes to governance regulation.Footnote 3 Philosophically, our analysis offers an intriguing contribution to organization theory and the study of complex and hybrid organizations. While compliance to codes is often considered ‘good enough’, in our hybrid setting these ‘ticking the box’ approaches often resulted in what West (1989) would label following ‘fixed recipes.’ Our analysis of bottom-up approaches in the creation of codes of governance, in contrast, takes ethical dilemmas as a starting point, rather than evading them by relying on fixed recipes. This more pragmatist approach seems to tally well with the institutional reality of hybrid organizations, because it seems to be an apt way of coping with governance issues whenever compliance does not seem ‘good enough’. While organizations and their bodies of governance tend to embrace fixed recipes in order to adhere to the imperatives of their shareholders, stakeholders or, in the case of cooperatives, their members, we saw that the new solutions found in our empirical context arose out of pragmatism or, in the words of Dewey (1930): ‘What works’. This is an important insight for the study of good governance because it challenges the important assumption that, in order to steer organizations towards prosocial behavior, rule-based ethics is a promising approach. It is therefore no surprise that cooperatives are known to be notoriously late with the integration of CSR schemes, which often draw on the same deontological perspectives as the codes of governance that have proven dysfunctional (Pollach et al., 2012; Seguí Mas et al., 2016). For the literature on alternative organizations, especially those where agency models do not resemble the typical separation between ownership and control, deinstitutionalization and a pragmatic view on what works, rather than institutionalization, may be a fruitful strategy to steer organizations towards good governance.

Practical Implications: Recommendations for Effective Code Creation for Hybrid Organizations

Do Not Prioritize

The analysis of our data suggests that institutional hybridity should be taken seriously throughout the creation process to avoid two impending dangers. On the one hand, neglecting one or more logics in a governance code leads to low uptake or superficial implementation when not taken seriously, and may effectively destabilize hybrid organizations rather than contributing to their good governance (Sundaramurthy & Lewis, 2003). On the other hand, failure to explicitly recognize the challenges associated with organizational hybridity should be considered a missed opportunity, as scholars have noted that awareness of paradoxical tensions results in novel, innovative practices contributing to organizational success and sustainability (Smith et al., 2015). Taking hybridity seriously, both during and after the code creation process, highlights the need for co-constructed, principle-based governance codes that can be implemented through dialog.

Do Not Prescribe

Governance codes may target a group of hybrid organizations that share adherence to the same institutional logics. Our results clearly show that their operationalization, and thus the challenges to which they give rise, are specific to the organization and its context. Indeed, institutional logics guide the conceptualization and justification of organizational practices for organizations and members (Thornton et al., 2005). However, they explain not only existing practices, but also the adoption of new ones, through a path-dependent process (Shipilov et al., 2010). Consequently, if governance codes are to address the governance challenges of either contestation or estrangement, their relevance will depend on the specific characteristics and context of the organization. Rather than providing rigid prescriptions for practices, this highlights the need for active, deliberative implementation at the organizational level. This approach to implementation will not only minimize the risk of institutional decoupling (Bromley & Powell, 2012), but will also leave room for increased adoption of heterogeneous practices, spurred by institutional hybridity (Ocasio & Radoynovska, 2016).

Do Not Stop Learning

Institutional hybridity drives organizations to experiment and develop innovative practices (Ocasio & Radoynovska, 2016). Consequently, collective learning within the sector should be actively supported. Ways in which organizations prioritize institutional logics may differ, yet they can still benefit from insights into how conflicting institutional demands are reconciled in practice in similar organizations. As indicated by our results, this is especially relevant for organizations in contexts of estranging hybridity. Collective learning strategies can also be used to strike a balance between innovating in the face of governance challenges, while sharing the risk of new governance failures (Canales, 2011). Therefore, a key role for institutional prescriptions for hybrid organizations lies in dynamically supporting identification, dissemination, and improvement of practices adopted by individuals and hybrid organizations that face similar constellations of institutional logics.

Limitations and Future Research

This study has some limitations that suggest paths for future research. First, despite our careful theoretical sampling, generalization of our findings to the entire cooperative sector, or even hybrid organizations in general, should be undertaken cautiously. Grounded theory is, by nature, not intended for generalizing findings to a broader population per se, but for uncovering actions/interactions under specific conditions (Corbin & Strauss, 1990). Nevertheless, there are several opportunities for future research to extend our findings. First, while our study sheds new light on the dangers of specific ethical underpinnings of codes, future research could more closely investigate different ethical stands and their relationship with institutional arrangements to mitigate governance failure (Chakrabarty & Erin Bass, 2015). Second, our conceptual distinction between adapted and co-constructed governance codes opens the way to broader quantitative studies. In particular, these studies could more systematically assess their uptake and effectiveness. Third, future research questions might apply similar research designs to explore the relevance of these code creation mechanisms and governance code roles in the context of other alternative organizations, such as meta-organizations (Gulati et al., 2012), non-profit organizations (Wellens & Jegers, 2014), social enterprises (Mair & Martí, 2006) as well as family firms and SMEs (Lee & Rogoff, 1996).Footnote 4 This research could also help to understand boundary conditions of our findings and shed light on which tensions are inherent to the cooperative model. Also, while the purpose of any exploratory process study is not to provide generalizable results, but rather to provide a consistent and internally valid representation of how governance prescriptions, in our cases, fail or succeed, we like to point out that the mechanisms we have revealed are non-exhaustive and that scholars could look for alternative processes of code creation by, for instance, also looking into cooperative movements filled with so-called “weak” cooperatives that do not strictly adhere to ICA principles. In addition, future research could seek to extend our finding by studying a governance code creation process as it unfolds. Last, future research could study and try to understand the potential of different forms of institutional prescriptions and good governance guidelines for organizational hybridity.