The Logic behind the Internal Governance of Sweden ’ s Largest Agricultural Cooperatives

: This study comprises a qualitative analysis of the governance structures within the two largest agricultural cooperatives in Sweden, both of which have large and heterogeneous memberships, as well as sizeable and complex business operations. Interviews were conducted with district council members and representatives from the boards of directors. The data from both case cooperatives indicate a genuinely traditional democratic member governance structure. There is a great deal of social capital in the governance system. A tentative explanation of the member involvement is that the cooperatives ’ equity capital is owned by the individual members, and the members are free to sell and buy their shares at a market rate. The members are satisfied with the return on the capital they have invested in the cooperatives. They receive bonus shares and dividends above what most members would get in other investments. A conclusion is that very large cooperatives may succeed well with a member-democratic governance system of the traditional cooperative type, which in this case is probably due to the members’ satisfaction with the individualized ownership of the cooperatives.


Introduction
The study presents unique cases about the ways in which large agricultural cooperatives could contribute to sustainability. Since the two investigated cooperatives introduced individualized ownership structures, more social capital within the memberships have strengthened the member democracy, whereby the cooperatives have added to the economic, social, cultural and environmental sustainability for their members, as well as for non-members, the countryside and society at large [1]. The investigated cooperatives play a major role in the Swedish agricultural sector, and they-as well as their predecessors-have been important throughout the history of Swedish cooperatives [2][3][4].
This study investigates the internal governance of the two largest agricultural cooperatives in Sweden. Both of them have diversified operations and heterogeneous memberships. "The internal governance refers to the decision-making processes, the role of different governing bodies, and the allocation of control rights to management (and the agency problems that goes with delegation of decision rights)" [5] p. 6.
The most often mentioned governing bodies in cooperative governance are a general assembly and a board of directors [6]. A general assembly is where the members gather to decide about their cooperative's strategic issues and to monitor the performance of the chief executive officer (CEO), who is appointed by the board to execute the strategic plans [7,8]. However, in a large cooperative, the general assembly consists of delegates, who are elected at district meetings. Each district is headed by a district council, which consists of members who are elected at district meetings [9,10]. As the district councils elect delegates to the general assembly, they constitute links between the membership and the board of directors. Except for electing representatives to the general assembly, the district councils update the board about rank-and-file members' opinions, and inform the membership about the board's decisions and the cooperative's performance. By focusing on the boards of directors and the district councils, the present study comprises the entire governance chain, from the members to the district councils, the general assembly, the board of directors, and finally the CEO.
Like other business organizations, cooperatives are hierarchically structured. Cooperatives, however, consist of two hierarchies: one for the membership organization (the society), and another one for business operations (the firm). The present study concerns only the membership organization. Power is successively delegated from the lower echelon of the membership organization to the next one up. A good governance structure is designed so that the higher echelons work in the interests of the lower ones [11]. Cooperative governance structures have become more varied during recent decades. Due to increasing vertical and horizontal integration, many cooperatives have complex operations, and the memberships have become large and heterogeneous. Many cooperatives have thus introduced non-traditional governance models [12][13][14].
The non-traditional governance models have gained the attention of researchers. A span of governance models have been explained in terms of members' property rights and control rights [15]. Authors have identified different governance models among European agricultural cooperatives [6] and on a global scale [16]. These studies have an aggregate approach, and are therefore not fully compatible with studies which are based on data from within the cooperatives; as such, they study the specific governance, for example in New Zealand [17,18], member reactions to changing governance structures [11], member involvement in governance [19,20], the influence of non-member directors, and, not the least, the relations of CEOs with their board of directors [5,6,16,[21][22][23][24].
When it comes to explaining why cooperatives have chosen one governance model or the other, the state of knowledge is, however, unclear, which is the reason for the present study. Some researchers suggest that cooperatives with complex and large business operations tend be governed by externally-recruited directors, because the members are not sufficiently qualified [6,8,16,25,26]. Other researchers have suggested that the cooperatives' financial structure affects the government system, because external financiers will demand formal influence [15,24]. Surveys among US cooperatives, however, do not support the hypothesis about a relationship between the governance structure and the financial structure [27,28].
The present study challenges the claim of an alignment between the cooperatives' governance structures and their financial structures. The two investigated cooperatives adhere strictly to the democratic member principles, which are typical for traditional cooperatives. The cooperatives' strategies concerning market-based and differentiated pricing, as well as individualized member capital, however, do not follow the traditional cooperative principles. The explanation is that, owing to the democratic member governance policy, the cooperatives receive member support for untraditional marketing and financial strategies, and these are instrumental in the provision of members' benefits.
The aim of this study is to explore the logic behind the governance structure of cooperatives, for which the two largest Swedish agricultural cooperatives provide the empirical basis. The study adds to the theoretical understanding about the factors that affect cooperatives' choices of governance structures.
For institutional reasons, the two investigated cooperatives are not representative for agricultural cooperatives at large, in Sweden or elsewhere. The legislation on cooperatives varies among countries and over time. Swedish law does not set any limits on how much money cooperatives can transfer to their members and still enjoy single taxation. This allows transferability for equity, and the cooperatives may have a variety of financial instruments. In spite of these peculiarities, the findings of this study may have a general and theoretical value, as they indicate the ways in which a cooperative's choice of government model is related to individualized member ownership and the cooperatives' business relations with members, while previous studies have focused on governance structures in relation to other financial structures. Section 2 presents a conceptual framework in which it is posited that the agency problems that are immanent in the governance of large cooperatives can be understood in terms of measures for the support of social capital. Section 3 describes the two case cooperatives. Section 4 gives an account of the data collection procedure. In Section 5, the observations from a series of interviews are summarized, while Section 6 comprises a discussion of the ways in which the data can be interpreted in terms of social capital. Section 7 presents conclusions about the links between governance, ownership, and marketing principles.

The Costs of Internal Governance
Some cooperatives have small and homogeneous memberships, and simple, member-related operations. The entire membership can gather at general assembly meetings, and the members make decisions on the cooperative's strategies. Thus, the members have low costs for running their cooperative. Personal acquaintances mean trust, and the members will have similar notions about their cooperative's mission. These conditions were prevalent when practically all existing cooperatives were young, but they can still be found in many small cooperatives [29]. However, it became necessary to have a leader, as many of the small cooperatives expanded. This person was often a trusted member or a farmer's son. This entailed increasing, but still relatively low, monitoring costs. Communication was simple through personal contacts. The interdependence between members, as well as between members and leadership, fostered social capital.
As time passed, intensified competition and a quest for low processing costs through large-scale operations gave rise to many mergers. Thus, cooperative memberships have grown to thousands, or tens of thousands. If members of a large and heterogeneous membership are to unite about conditions that everybody agrees on, there will be high collective decision-making costs. "Collective decisionmaking costs arise with the adoption of costly processes to address patron interest heterogeneity and/or inefficient decisions that fail to maximize the aggregate welfare of the owners as a group" [16], p.21.
Hansmann [30] suggests the delegation of authority to committees as a way of reducing collective decision-making costs. As the delegation of power in the member hierarchies has come to comprise several echelons, the distance between members and leadership has increased. Instead of collective decision-making costs, cooperatives have higher democracy costs in the form of meetings and information letters, etc. [31,32]. However, these cost increases are outweighed by the production cost savings, owing to economies of scale. In the same manner, more efficient production has compensated for the higher cost to members of monitoring leadership in the form of incentive schemes, auditing, policy documents, annual reports, and meetings at different organizational echelons, etc.
When a cooperative's operations become more complex and the membership organization is split into several layers, there will also be agency costs. Agency costs are a consequence of information asymmetry, which arises when somebody, due to their superior knowledge, is able to reap undulyhigh benefits [33]. The principals, i.e., the members of a cooperative, will suffer as the agent, i.e., leadership, extracts part of the residuals that the members are entitled to in their role as owners and users.
Agency theory is based on the assumption of opportunism, i.e., self-interest seeking with guile. Some humans may sometimes seek advantages for themselves, even to the extent that they are not honest with others. Due to information asymmetry, the leaders of a hierarchically-structured organization are in a position to deceive others [34]. CEOs are in a special position to act opportunistically, as they are tasked with the execution of the decisions made by the boards, and a board may have difficulties in monitoring the CEO. There may also be opportunistic behavior in other relationships within the membership hierarchy. Even elected representatives can be deceptive.

Agency Costs and Social Capital
Cooperatives with large and heterogeneous memberships are more likely to suffer from agency problems, and so are cooperatives with complex businesses [35]. Under these conditions, the amount of social capital within the membership may be so low that the members do not make sufficient efforts to govern the cooperative [36].
Cooperatives have been characterized as being based on the existence of social capital within the membership [37,38]. For internal governance to result in the coordination of activities and a good allocation of resources, there is a need for social capital among members, elected representatives at different levels, and management. [39] p. 6, specify three attributes of social capital: "the fabric of trust, shared values and understanding that allows diverse participants to work together towards collective outcomes and common goals".
Trust: interpersonal relationships can lead to trust, reciprocity, and solidarity [40,41]. Trust is "the willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party" [42] p.712. However, the trust evaporates if a trusted partner is found to be dishonest. Therefore, trust is more likely in long-term relationships, and when the parties are mutually dependent on each other. The relationships between members and elected representatives at different levels of the membership hierarchy reduce the information asymmetry of the organization, and thus the amount of trust increases. Elected representatives are often keen to have a good reputation.
Shared values: elected representatives have the task of facilitating information exchange, which contributes to interest alignment, mutual relations, values, norms, a motivational structure, and social skills, all of which facilitate communication. Researchers have formulated so-called 'cooperative values', claiming that the cooperative principles originate from a specific set of human values, such as equality, fairness, and democracy [43][44][45]. These values are supported by legislation, statutes, contracts, information, and other cooperative phenomena, but they cannot, in themselves, guarantee low agency costs. More interest alignment is achieved through shared values, and more common norms are fulfilled.
Understanding: interaction fosters social norms, which contribute to cohesion. Social exchanges between equals stimulate a feeling of community. Because cooperative legislation and statutes stress the equal and fair treatment of members, there is the probability that mutual understanding will evolve. Networking facilitates collaboration and stimulates learning [40,46]. Networks can be characterized by bonding and bridging social capital. While bonding social capital refers to social networks of closely related people-such as family, close friends, and neighbors-bridging social capital exists among loose friendships and colleagues [47]. Bonding social capital provides personal support and access to resources, and is crucial in the creation and nurturing of community organizations [10,47].
In short, social capital affects people's involvement in collective activities, such as the governance of cooperatives [48,49]. Reciprocity is expected in well-functioning governance processes, i.e., it provides a service to others with the expectation that the kindness will be returned [41]. A key concept is legitimacy, i.e., that the membership thinks that the leaders govern the cooperative in a way that is in the membership's interest.

Sweden's Largest Agricultural Cooperatives
This study analyzes the governance of the two largest Swedish agricultural cooperatives, namely Lantmännen, which is a grain marketing and farm input cooperative, and Södra Skogsägarna (hereinafter 'Södra'), which is a forestry cooperative. In Sweden, forestry cooperatives are considered to be a branch of agricultural cooperatives because about half of the country's productive forestland is owned by farmers and other individuals [50].
The number of Swedish agricultural cooperatives has fallen to approximately twenty due to mergers and acquisitions [3]. This means that there are both large and small cooperatives; some of these are market-dominant, while others apply a niche strategy [29]. The two case cooperatives dominate many of the markets in which they operate. Table 1 presents the cooperatives' net sales, equity ratios, patronage refunds, dividends to members, and other details.
Lantmännen has about 25,000 members, while Södra has about 50,000 members. Lantmännen allows active agricultural producers throughout Sweden to become members. Södra requires that a member owns forestland in the southern part of the country. Some Lantmännen members in southern Sweden own forestland in addition to their agricultural land, and therefore also belong to Södra. Even though both cooperatives have extensive international operations, there are no members from other countries.
The case cooperatives have an open membership policy in accordance with traditional cooperative principles and Swedish legislation, i.e., they have equal voting rights and delivery rights for members. However, both cooperatives have traits of being non-traditional, or hybrid, cooperatives because their members have-to a large extent-individualized ownership, and thus the cooperatives are partly profit-seeking. This is also supported in that the cooperatives' business relations with their members are strictly based on market considerations, while solidarity in the form of cross-subsidization and other traditional cooperative principles has been abandoned.
The introduction of individualized ownership in the case cooperatives took place over some years, starting in about 2000 (Södra) and 2010 (Lantmännen). One reason was that the farmers are becoming fewer, have increasingly large farming operations, have large investments, and are professional, all of which indicate an trend towards a focus on economic benefits and an individualistic mentality [4,51]. In order to keep member involvement, business volumes, and equity, the cooperatives adapted to the emerging member requirements. The benefits of the individualized ownership are confirmed by surveys that the cooperatives' member relations offices conduct regularly, and by surveys conducted by researchers [1,52,53].
The individualized ownership structure of the investigated cooperatives is congruent with the definition of cooperatives: "In a cooperative, the user is the focal point, with the direct status of user, owner, and control vested in the same individual" [54] p. 85. This definition does not claim anything about collective versus individual ownership, nor do other conceptualizations [30]. Nevertheless, other Swedish cooperatives have kept their traditional financial structure, with non-tradable shares and a large amount of unallocated equity. They would not have a large enough amount of tradable shares to be attractive to any bank to function as a market maker, and none of them have such a stable financial status and such good performance as the two case cooperatives.
The next two subsections give an overview of the two case cooperatives' development, business operations, ownership, and member composition. The presentation of the governance structures is merged in the subsequent subsection because the two cooperatives are similar in this respect.

Lantmännen (Source: [55])
Development: Lantmännen is the result of a large number of mergers among smaller cooperatives; the oldest was established in 1880. The present cooperative was created in 2001 through an almost-nationwide merger of nine regional cooperatives, creating a new nationwide cooperative. It is one of the largest agri-business firms in Northern Europe. The cooperative is continually expanding its operations, mainly in foreign markets. Lantmännen has production plants in more than twenty countries.
Business operations: business operations are split into five divisions. The members have business relations only with 'Division Agriculture', which represents 27% of the cooperative's sales. This division purchases all types of crops and sells all types of farm inputs, such as seed, pesticides, fertilizers, and machinery. In 2016, 41% of the grain was exported. The division operates in accordance with traditional cooperative principles. Given the heterogeneity of the membership, the cooperative is keen on avoiding cross-subsidization, and thus the prices that the cooperative offers its members are based on market considerations. All of the profits that the division makes are reimbursed to the members in the form of patronage refunds, the issuance of bonus shares, dividends on common shares, and bonus shares. The bonus shares are freely tradable among the membership at market prices.
The other four divisions are profit-seeking. The largest one, 'Division Food' (36%), processes member and non-member grain into flour, bread and cakes, pasta, cereals, and other products. These products are sold under an array of well-known brand names in many markets. The other divisions are 'Division Machinery' (29%), Division Energy (bio-ethanol and alcohol) (7%), and 'Division Real Estate' (1%).
Ownership: members are obliged to own common shares in direct proportion to their volume of business with the cooperative, although payment is made over many years through deductions from the payments from the cooperative. In order to finance the four profit-seeking divisions, Lantmännen sells preferred shares to members. These are freely tradable between members at a market rate. At least 40% of the profits made in the profit-seeking divisions are paid to members as dividends on their preferred shares. The capital return of members is far higher than they could get in most other investments ( Table 1). The remaining 60% goes to the unallocated funds, which are later converted into bonus shares.
Membership heterogeneity: the Lantmännen membership is quite heterogeneous. Its large geographical spread means that its members have different production conditions. The members may have all types of agricultural production. Many are part-time farmers. Some members produce grain, which the cooperative converts into animal fodder to be sold to other members. Grainproducers and animal breeders are likely to sell their products and buy farm inputs regularly, but their solidarity cannot be counted on. While most members are interested mainly in their role as patrons, others want returns on their investments in the cooperative.
3.2.2. Södra (Source: [56]) Development: Södra was established in 1938, but since then, a few neighboring forestry cooperatives have been included, whereby Södra's operations have come to comprise the most productive forestland in Sweden. The cooperative has-during the last five to ten yearssuccessively developed a specific model for individualized member ownership, implying that members are incentivized to invest in the cooperative.
Business operations: the cooperative's business operations cover the entire process, from planting new trees to processing the various components of trees into marketable products. The operations take place in four divisions. 'Sö dra Forests' is the member-related division. It buys timber from members and offers all sorts of forest management services to its members (21% of the net sales). 'Sö dra Wood' (30%) runs sawmills and processes sawn timber into other products, such as houses. 'Sö dra Cell' (42%) runs the paper pulp production. The remaining 7% consists of miscellaneous activities, such as wind mills. The cooperative has many processing plants in Sweden and some neighboring countries. Södra sell paper pulp and sawn timber to other processors. With about 57% of its production volume being sold abroad, Södra is one of the world's largest exporters of paper pulp and sawn timber.
The processing plants represent large investments. The paper pulp mills have a process production, which means that they run close to their capacity ceiling. This requires an even inflow of timber, which is attained because Södra uses the price parameter when buying timber from its members [1]. Södra has a policy of evening out the timber price over the years, as a high price during a low business cycle would result in volumes that cannot be handled.
Ownership: Södra's pricing policy results in profits. About half of the profits are paid to members as capital returns and patronage refunds, while the remainder goes to unallocated funds. The money is paid to the members in a way that creates a balance between the members' willingness to invest and to deliver. The members receive patronage refunds. It is more important that Södra has a specific set of rules regarding common shares, bonus shares, and preferred shares. When delivering timber, the members receive common shares when a withdrawal is made from the payment for the delivered timber. Södra has no other requirement for the member ownership of common shares.
Common shares are attractive for members not only because they get a dividend, but because they also receive annual bonus shares in proportion to their ownership of common shares, and the ownership of bonus shares is also rewarded with dividends. The bonus shares emanate from the cooperative redistributing a part of its unallocated funds to its members. In addition to common shares and bonus shares, the members have the possibility to buy preferred shares, which are subject to dividends as well.
Members can freely buy and sell all three classes of shares on an open market. The market for shares and the high return on member investments are core elements in Södra's business model, because the members are interested in receiving a high return on their investments. The members decide for themselves how much they want to invest in Södra. Thus, the member ownership of shares is basically voluntary. Södra's ownership model implies that the cooperative works according to profit-seeking principles.
By catering to market mechanisms both for the raw product market and the capital market, Södra can convey more money to its members than what would be possible if all of the traditional cooperative principles were adhered to, i.e., paying a high price for timber and paying a low return on member investments. Surveys among Södra's members indicate high scores on member satisfaction [52,53].
Membership heterogeneity: even though all of Södra's business operations are related to wood, the membership is heterogeneous. A growing share of the members do not live at their properties. Some members mainly focus on their patron role, while others consider their investor role to be more important. Some members disapprove of Södra purchasing timber from non-members and its foreign investments. Forest owners' business with Södra may be quite irregular because they are in a position to postpone their deliveries for years, and likewise, the services that the cooperative provides can be postponed for several years. Figure 1 applies to both of the investigated cooperatives, since they follow the same legislation on cooperatives and are subject to the same set of traditions. The similarity between the cooperatives may also depend on the fact that both are large and have integrated a great deal forwards into processing. The arrows in Figure 1 indicate that there are many agency relationships involved. Table  2 shows the differences in the number of representatives in the various assemblies.

Governance Structure
The Swedish law on cooperatives prescribes a traditional governance structure. The general assembly is the highest decision-making body in a cooperative. The general assembly in larger cooperatives consists of delegates, who are elected by district assemblies to which all members are invited [10]. When this two-tier system was introduced to substitute the former three-tier systemwith local, regional, and apex levels-the cooperatives obtained substantially lower democracy costs, as the number of elected representatives was reduced [57].
The board members are elected by the general assembly. Both of the investigated cooperatives elect a few non-member experts as directors. Likewise, the CEO takes part in many board meetings, and the top management in both cooperatives has a good reputation in the Swedish business world.  As can be seen in Figure 1, the members belong to geographically-defined districts. A district council is elected at the annual district assembly meetings. The chairpersons of the district councils constitute the general assembly, which meets once per year. The general assembly elects the board of directors. Swedish legislation does not require advisory councils, but such have existed for an extended period in both cooperatives. All of the advisory council members are elected by the general assembly. Their task is to control and assess the work of the board and the CEO. An advisory council, which has no decision rights, meets a few times a year, together with the board members, the chairpersons of the district councils, a few other chairpersons from the largest districts, the chairperson of the general assembly's nomination committee, and the lay auditors [10].
Nomination committees are not required by law, but they exist whenever an election is to take place [58]. A committee proposes one candidate for each of the positions that are to be filled. The committee members are elected at the preceding year's assembly meeting from among the members.
The general assembly not only appoints an auditing bureau, but also elects lay auditors from among the members. A few members are tasked with the analysis of the work of the board and the CEO. This practice has been followed in all cooperatives for an extended period, but no legal requirements exist.

Data Collection
This study requires a qualitative approach in the form of case studies [59][60][61]. It was not possible to specify beforehand which data were needed in order to outline the governance of the two cooperatives. There was a need for two-way communication with follow-up questions. For that purpose, the data was collected through semi-structured interviews with key informants. The interviews were conducted either in person or by phone.
During the summer of 2016, the research team conducted interviews with eight delegates of Lantmännen's district councils and six delegates from Södra's district councils, mainly face-to-face at the interviewees' farms, throughout the cooperatives' operating areas. The interviewees were identified through a process of random sampling. During the spring and summer of 2017, telephone interviews were conducted with the chairperson, two other directors of Lantmännen, the head of member relations, and the CEO. The corresponding representatives were interviewed at Södra, but the chairperson declined due to limited time, and unfortunately, it was not possible to set a new date for an interview.
All of the interviews were recorded. Written summaries were sent to the respondents with a request for eventual corrections. All of the interviewees responded. Most of them had no corrections to the summaries. The few comments were of a clarifying nature.
The semi-structured interviews were conducted with one set of questions for each interview with the district council delegates in both cooperatives, and another set of questions for the interviews with all of the board representatives in both cooperatives. Because the interviews focused on ways of handling agency problems, there were questions about information exchanges, meetings, responsibilities, member involvement, and similar variables. Before the interviews, the authors compiled question guides for the different categories of interviewees. The questions to the district council members were categorized in the following classes: (1) in relation to the members; (2) in relation to the board; (3) information channels and meeting arenas; and (4) the district council's role. Likewise, the questions concerning the boards of directors contained issues in the following categories: (1) in relation to the district council chairpersons; (2) in relation to the management; (3) information channels and meeting arenas; and (4) the role of the board. The following section's account of the interview results are categorized in the same way, while the results in Section 6 are interpreted in terms of social capital theory.

Lantmännen's District Councils
Relation to the members, the district council delegates regretted that the members rarely contacted them. The reasons mentioned were that the districts had become large, both in geographical terms and in the number of members. Some interviewees believed that the low frequency of the contact from the members was because the members perceived a long distance between them and the district council delegates. There was less commitment within heterogeneous districts. Likewise, Lantmännen's policy of redistributing profits made the members satisfied and less active.
Several respondents said that members read the annual reports as well as other written statements. The members were fairly engaged in Lantmännen's activities. The interviewees, however, raised a concern when it comes to the member attendance at the meetings arranged by the district council. Production-oriented meetings and farm visits were better-attended than meetings about strategic, economic, or 'cooperative' issues.
Some interviewees said that it was difficult to decide which member sub-group to listen to, but "you learn who to listen to after a while". They stressed that it was central to always "hear members out, but use your common sense when it comes to which opinions to transmit to the board". They said that it was important to develop trust among members, as well as to be good at adjusting the way in which the communication is carried out.
Relation to the board: a core characteristic of being a district council delegate, mentioned by the interviewees, is: "you have to consider it as a vocation". Obtaining trust for the district council member among the Lantmännen's board of directors was crucial. The importance of being able to discuss issues with the board members was also stressed. It should be kept in mind that the firm is a cooperative, which means that short-term profits are not the only focus.
Information channels and meeting arenas: the interviewees believed that members read the newsletter, since the issues brought up there were the same ones that members raised when meeting district council delegates. One interviewee said: "In a way [the newsletter] is Lantmännen". The interviewees said that few members took the time to read the entire annual report: the main reasons given were that the report is too verbose, with more than 100 pages. However, the summary of the annual report was read by many members.
The interviewees considered the website to be an important source of information, as arenas for digital meetings will be, but the respondents stressed the importance of continuously updating all of the websites. Besides being a source of information, the Internet will be valuable for "virtual meetings".
The respondents felt that offering courses and information meetings to members was important. Over the years, Lantmännen has offered both courses and meetings. These have focused on being an owner of a cooperative, how the cooperative's operations are run, and the benefits for its members. Every fall, Lantmännen arranges meetings about the role of being a member of a cooperative.
Most district council delegates stressed the importance of the advisory council. "It is a very important arena for discussions and the dissemination of information", said one respondent. The discussions of the advisory council improved once they were made confidential. Before confidentiality was introduced, the delegates were concerned about what issues discussed by the advisory council could be mentioned to members. One interviewee described how he handled this: "Things I am allowed to talk to members about, I write down, while the rest I keep in my head".
The district councils' role: district council delegates have a twofold task: forwarding member opinions to the board, and explaining the decisions made by the board to the members and discussing them, mainly at meetings but, in some cases, also in member leaflets. Some of the interviewees mentioned that it is important to reach a situation where "both members and directors trust you, and be able to alternate between discussing with members and directors". The Lantmännen District Council delegates think that there is a relatively short distance between the members and the board. Some members even contact the directors instead of turning to district council delegates.

Södra's District Councils
Relation to members: all of the members have their forest estates in the Södra area, but an increasing number do not live on these estates. This group is often less involved in forestry. This leads to problems when district meetings and activities are arranged, since most non-residential members live far away. One way to improve attendance is to announce that an important strategic issue will be discussed at a district meeting, and another method is to announce that one of the directors will be attending a meeting.
The district council delegates struggle with the low commitment of some member groups, as well as a lack of knowledge about how cooperatives work and the markets in which Södra acts. Some cooperative vocabulary is not understood by some members. Dealing with this requires good communication skills and trust building.
The district council delegates said that members seldom contact them, and if they do, the contact is usually initiated by a specific incident, for example, "when the CEO was sacked", as one of the respondents said. The poor member contact was explained by the merging of the districts, which meant that some members had to travel far to attend a member meeting. The number of district council delegates is decreasing, but the requirements put on these, in terms of interacting with members, have remained the same, or are increasing.
One district council delegate said that one reason why members tend not to come to member meetings is that they think the agenda mainly concerns formal issues and information from the board, and listening to this is a waste of time. They would instead prefer to be informed through the website, for instance. This delegate said that the members' interest in attending meetings would increase if the board would "present a 'real problem' to be discussed at the meeting".
Another issue mentioned in the interviews was the average age of the members, which was high, and the elderly members were more likely to attend the meetings more often than younger members. This led to a tendency to "stick with old ideas and that younger members do not learn about the cooperative". Biannually, so-called 'Owner Meetings' are arranged, where 8-10 members are invited. These meetings focus on issues related to the ownership of Södra. Those who attend these meetings are usually satisfied, the respondents said, but a problem is that only a few members attend these meetings.
Relation to the board: the interviewees described the dual role of district council delegates as both "taking care of the Södra brand, and offering 'member service'". They appreciated the latter role as the most interesting. Some expressed the idea that some of the duties of the former role could be handed over to the officials of the cooperative.
Information channels and meeting arenas: the newsletters published by the headquarters are read by members, while the annual reports are not. As at Lantmännen, even Södra District Council delegates admitted that they did not read the entire annual report. The district councils' newsletters about the events in the district were read by most members. The courses offered to members in the district-often with forestry production content-were appreciated by many members, according to the respondents.
The Södra Advisory Council meets three or four times a year. It is a forum for district council chairpersons and board members. The interviewees said that the advisory council meetings are very valuable; they decrease the distance between the districts and the board, and the issues discussed are interesting. One of the respondents, who had been a district council delegate for Lantmännen and attended the meetings of the Lantmännen Advisory Council, said: "it feels as if you are closer to the board of Södra", which he believed could be a result of the difference in the geographic size and the scope of the two cooperatives.
The district councils' role: each district council delegate has a certain area of responsibility, for example, issues related to safety, policy, and rules and regulations. The idea is that a district council should be prepared to arrange meetings and courses when the need appears. The district councils entail "monitoring local business policy and organizing training courses and member activities". The delegates "play a key role in dialogues with individual members".
Because a district council is the link between the members and the board, trust is essential. One problem mentioned by the interviewees is that the average age in the district councils is fairly high. Furthermore, because some members live far away from Södra's area, there are problems relating to meeting attendance and member commitment.

Lantmännen's Board of Directors
Relation to district council chairpersons: the directors highly appreciated meeting the district council chairs at the advisory council meetings. These meetings gave the directors the possibility to discuss strategic issues with the district council chairpersons, as well as to listen to the members' opinions presented by the district council chairpersons. Some people also met a few times each year to discuss issues such as "being an owner of Lantmännen", and the strategic and financial issues.
Relation to management: all of the interviewees stressed the importance of keeping the roles of the board and the CEO separate: "The board should focus on strategic issues, while the CEO focuses on operational issues". Because the CEO "controls all the resources, it is important that the board makes the strategic decisions". The chairperson said that the division of roles was clear and decided upon, but "you have to constantly repeat it-especially since there are new board directors as well as new staff in the management team".
Lantmännen has a recurring item at the board meetings, labelled 'information from the CEO,' where the CEO discusses "how the board thinks he is to act in different ongoing issues, which is valuable both for him and the board members". The chairperson and CEO also have recurrent followup meetings between the board meetings, in order to discuss current problems and developments.
The CEO is evaluated throughout the year. The CEO is asked to present quarterly financial reports of about 40 pages at the board meetings, focusing on financial issues, as well as "a monthly report, focusing on the results generated". Once a year, the CEO's performance is analyzed more deeply during a board meeting without the CEO present. The result of this CEO evaluation is discussed by the chairperson and the CEO.
Information channels and arenas: Lantmännen publishes a traditional annual report as well as a report on Social Corporate Responsibility. The reports are offered to the members both via the website and via e-mail. These extensive publications are summarized in a short publication, with a focus on the financial data, and are offered to members on the website as well.
All of the interviewees said that the newsletters are an important information channel. The chairperson and the CEO write about current issues and summarize recent business developments. As a complement to the newsletter, Lantmännen has a member portal at its website, as well as a Facebook group. This is mainly in order to reach younger members, one respondent said. Lantmännen is convinced that there will be a need for more digital communication channels, and the cooperative is developing such channels.
The Lantmännen Advisory Council is seen by the directors as a valuable meeting arena, especially after these discussions became confidential. One director, who also sits on the boards of other firms, summarized the main difference between these boards and that of Lantmännen: "the anchoring process, without a doubt!" He said that the anchoring process, including the discussion of strategic plans in the advisory council, has an important role, since it "gives members the possibility to understand the logic behind the strategy discussed also in cases when members do not agree".
The member newsletters are published on the external website about four times per year, and usually cover 24 pages. Such newsletters treat recent market events, explain cooperative vocabulary, and present internal facts and figures. In some districts, the newsletters are distributed by mail to the members. In order to simplify for district council delegates, there is a policy that, when major events are about to take place, the district council delegates should be informed at least 24 h before the members are informed.
The role of the board: in 2016, the Lantmännen Board of Directors consisted of nine directors and three employee representatives. The deputy chairperson said that he "has always had the opinion that a board should not have more than seven directors", but that opinion has now changed. He used the argument that the risk for "freeriding directors" is higher in a large board, but now he thinks that this risk can be balanced by the possibility of having a board that has the required competences and time to fulfil the task of a director. Hence, his opinion today is that nine directors are necessary for Lantmännen. The secretary of the board is the Chief Legal Officer (CLO), because it is convenient to have a person with legal knowledge present at the board meetings. Lantmännen has a work group that prepares the agenda for the board meetings. The group consists of the chairperson, the deputy chairperson, the secretary, the CEO and the Chief Financial Officer (CFO). There are 13 board meetings per year.
The deputy chairperson said that the experience of the two external directors is very beneficial for the board's work. The external directors must, however, be aware of the special characteristics of a cooperative, and must understand that close contact with members is a necessary aspect of running a cooperative.
All of the documents needed at the board meetings are to be downloaded from the website and read. One director said that all of the documents must reach all of the board members, and the documents must not be spread outside the board. One goal is to introduce a similar web-based system for documents to be read in preparation for district council meetings.

Södra's Board of Directors
Relation to district council chairpersons: the Södra Advisory Council comprises 36 district chairpersons, as well as the directors. This arena is highly valued by the directors. One of them said that the advisory council is a well-functioning forum for discussions, as it supports the district council delegates in their contact with the members. He concluded by saying: "without the advisory council, frustration and irritation among members would easily arise and lead to a 'we and them' mentality, which might threaten cooperative collaboration".
Relation to management: Södra's respondents stressed the importance of keeping the roles of the board and the CEO separate. They mentioned that it is necessary to occasionally discuss this and decide: "where the boundary between the board and CEO is", as the deputy chairperson said.
The CEO said: "a good CEO may offset a bad board, but a good board cannot offset a bad CEO", concluding that the ambition is to have both a good board and a good CEO. If the chairperson and the CEO have a good relationship, the situation is optimal, he concluded. He also stressed the importance of the CEO knowing what is unique for cooperative governance. The chairperson and the CEO have frequent discussions, which is a key component for reaching a point where the board contributes to successful development. The CEO is evaluated yearly in a board meeting where he is not present. In addition to this, the CEO is sometimes evaluated "in a deeper way", but the interviewees thought that it is usually sufficient to evaluate the CEO through the ongoing discussions between him and the chairperson, and through one yearly evaluation by the board.
Information channels and arenas: the annual reports and the Corporate Social Responsibility reports are accessible through Södra's website, and so are the member newsletters. All of the interviewees said that digital information channels-as well as meeting arenas-will be developed, partly because the members require it, and partly in order to handle the members who live far from their forest estates. Södra has a policy that at least one director and one representative from the management team are to take part in the annual district meetings.
The advisory council constitutes a valuable meeting arena, respondents said, but more informal meetings are arranged both between the district council delegates and the directors, and between the latter and the members. The member relations officer is also frequently contacted by the members. When they cannot give an answer, they forward the member concern either to the board or to the management team.
The role of the board: the Södra Board of Directors had twelve seats in 2016. Södra had one external director, who was appreciated by the vice chair, who was one of the interviewees. He said that external directors bring specialized knowledge and new ideas. In 2016, Södra's board had seven physical meetings, and four digital meetings. The deputy chairperson said that it is important to plan the meetings well, making sure that all of the necessary issues are dealt with annually. The Södra Member Relations Officer acts as the board secretary, which is valuable, the deputy chairperson said, since this simplifies the link both between the board and the members, and between the board and the business activities, because the CEO is also present at the board meetings.
Södra's current CEO has been on about 50 boards over the years. He thinks this experience is beneficial for his job. One benefit he mentioned is that he understands the two roles, i.e., the roles of a board and a CEO, and the fact that there is an information asymmetry between the board and the CEO. He appreciated that the chairperson and the CEO prepared the board meetings in a way that leads to "the relevant issues being discussed at the board meetings". The CEO said: "the relationship between CEO and chairperson is crucial". The board evaluates the CEO's performance every year at a meeting without the CEO being present. If adjustments are found to be necessary, the chairperson is responsible for communicating this to the CEO.

Discussion
This study supports the conclusions of Morfi, Nilsson, Hakelius and Karantininis (in press) [4] when these authors wrote about the essential role that social capital has in cooperative governance. A well-functioning government depends on the elected representatives' involvement. In accordance with Morfi, Nilsson and Ö sterberg (2018) [10], it was found that elected representatives' motivations are often various types of personal satisfaction, not least the possibility to interact with other representatives, and with rank-and-file members. The ability to build up positive relations to others is important.
The advisory councils play a crucial role for interactions to be fruitful throughout the membership hierarchy. The representatives from both the board of directors and the district councils appreciate their discussion during the advisory council meetings. The meetings are important for transferring knowledge about opinions at the grassroots and leadership levels. Participating in the advisory council meetings is also important for the participants' self-esteem.
For social interaction to be stimulating, the individuals must experience progress. The meetings and discussions not only result in decisions, but they are also a confirmation of the participants' achievements. As Eisenhardt [62] suggests, social confirmation is important when the targeted achievement cannot be measured, such as governance tasks, for which it is not possible to construct an effective incentive scheme. Both of the investigated cooperatives had measurable goals of a financial character-i.e., profits, dividends, and patronage refunds-but it is not possible to estimate the ways in which the individual elected representatives contribute to these goals.
Many members were not much involved in the governance of either cooperative. However, the members voluntarily decided whether to abstain from participation or not. The respondents' explanation was that member passivity is due to satisfaction and trust in the elected representatives, which is in accordance with Nilsson and Svendsen [20]. The data indicates that passivity occurs more often in certain member groups. Small and part-time members of Lantmännen have weak incentives to involve themselves. Non-resident forest owners do not care much about member democracy. In short, when small member groups are passive, there is no reason to worry, especially as these members account for a small share of the cooperatives' business activities.
This study supports Kronholm and Wästerlund [9] in their observation that there is a problem when elderly members, in particular, are active, and try to influence their cooperative to follow outdated policies. In order to handle this problem, the elected representatives must have social skills.
Information exchange is essential on all levels of the hierarchy, even though all communication is costly [8,32]. The amount of information transferred electronically has increased. The district council delegates communicate with their members about what takes place at higher levels of the hierarchy, whereby they create an understanding of the mission and the structure of the cooperative's business operations. The members learn the logic behind the cooperative's market strategies and financial arrangements. They may talk about their joint interests as well as issues concerning production.
The meetings that the district councils arrange for the members of the districts contribute to the creation of a sense of community, and even cohesion. Common norms are reinforced when members meet and fraternize [4,63]. The members have widespread agreement about their cooperative's financial and marketing orientation. The meetings have the major social significance of helping the members to become more acquainted with each other, thereby stimulating a sense of community.
While personal meetings are likely to have a higher impact on social community, there is also much written information within the democratic member system in the form of newsletters and websites. Electronic communication allows for reciprocal information exchange. The newsletters contain material written by the members.
The membership within a district is often too large to allow for personal acquaintances among everybody, but at least the district council chairpersons are well-known among the members, whereby there is potential to bridge social capital. It is likely that the chairpersons enjoy legitimacy and trust among the members, especially as the chairpersons have interesting information to convey. A certain degree of bonding social capital is likely to exist among the various district chairpersons, and between these chairpersons and the directors, as these have regular discussions during the meetings with the advisory council.
The heterogeneity of the memberships is reduced through information exchange. All of the members receive the same information from headquarters. The district information given to the members originates at the headquarters. The newsletters are perceived by the members to be informative, and they are appreciated. Their messages are based on market considerations, which the members understand.
Even though both cooperatives had professional leadership that secured financially sound results, their governance was characterized by a great deal of social capital [28]. Being a director is honorable, and provides a strong network with other important people. Directors enjoy a great deal of social capital, especially if a firm is successful. Leadership and especially the management team influence the discussion climate among members. The cooperative CEOs had a strong position, and the chairperson and the CEO prepared all of the board meetings jointly.
The nomination committees and the lay auditors gave legitimacy to the governance of the cooperatives. These units ensured the members that the board of directors and the district councils could be trusted. These bodies gave the members stronger control over their elected representatives, thus ensuring that they fulfilled their tasks.

Conclusions
This study explored the logic behind the governance structures of Sweden's two largest agricultural cooperatives. The analyses found that the democratic member system was characterized by a large amount of social capital throughout the membership hierarchy. The governance of the two cooperatives followed the principles of traditional cooperation. This may seem surprising, as the cooperatives' large and heterogeneous memberships and their complex business operations would, according to previous research, create member passivity.
The explanation is that the cooperatives have a hybrid character, in the sense that ownership is largely individualized, and the business relations with the members are market based. These hybrid attributes contributed to creating member involvement in the governance. These attributes help the cooperatives to support the members in their roles as both patrons and owners. The investigated cooperatives applied a policy of trading with their members on market conditions, although they also paid patronage refunds. This policy means that the cooperatives are profit-oriented. A large part of the profits is transferred to the members as bonus shares, and as dividends on common shares, bonus shares, and preferred shares. As shares are tradable among the members at a market rate, the members want their shares to have a high market price. Membership heterogeneity has less importance, since the members have an interest in high return on their capital. They are incentivized to invest and trade with their cooperatives, and sufficiently many members involve themselves in governance. The members also accept that centralized power is necessary in a cooperative with complex business operations.
The investigated cooperatives enjoy legitimacy because the members consider the cooperatives to support their economy. The elected representatives can see various problems, but they appreciate their tasks, and they are appreciated by the members. The existence of different organizational units contributes to social capital in the democratic member system, which reduces agency problems within the membership hierarchy.
A theoretical contribution of this study is the observation that cooperatives with large and heterogeneous memberships, and with sizeable and complex operations, do not necessarily have problems concerning their member governance. They may have a genuinely member-democratic governance structure as a consequence of individualized member ownership and market-set trading.