Control Rights, Governance, and the Costs of Ownership in Agricultural Cooperatives Fabio Chaddad Division of Applied Social Sciences, University of Missouri, Columbia, MO. E-mail: chaddadf@missouri.edu Constantine Iliopoulos Agricultural Economics Research Institute, Athens, Greece. E-mail: iliopoulosC@agreri.gr ABSTRACT Agricultural cooperatives have changed considerably in recent decades. In witnessing these structural changes, scholars have proffered analyses of nontraditional ownership models focusing on residual claim rights. However, crucial information on the allocation of control rights in cooperatives is missing. This study sheds light on alternative ownership-control models adopted by agricultural cooperatives in different regions across the world. In each of these models, we describe the allocation of formal control rights with a focus on decision management and decision control rights. We thus provide empirical evidence on the “separation of ownership and control” in agricultural cooperatives. We also analyze each of the governance models in terms of the associated ownership costs, including risk-bearing costs, the costs of controlling managers, and collective decision-making costs. In doing so, we are able to better understand the forces influencing the organizational efficiency of each cooperative model. [EconLit classification: C 2012 Wiley Periodicals, Inc. Q130].  1. INTRODUCTION Agricultural cooperatives have changed considerably in recent decades. More specifically, they have tinkered with and sometimes redesigned the traditional structure and adopted nontraditional models in response to changes in their competitive and institutional environments. In witnessing these structural changes, scholars have proffered descriptions and analyses of such nontraditional models (e.g., Cook, 1995; Cook & Iliopoulos, 2000; Hendrikse & Bijman, 2002; Holland & King, 2004; Nilsson, 1999). Building on property rights and agency theories of the firm, Chaddad and Cook (2004a) analyze the emergence of five nontraditional cooperative models from an ownership rights approach. However, these descriptions of alternative cooperative models are primarily based on residual claim rights. As a result, crucial information on the allocation of control rights in cooperatives is missing. In general, there is a dearth in the literature regarding the allocation of control rights and the structure of corporate governance mechanisms in farmer-owned cooperatives. A notable exception is the attempt of scholars from the Netherlands to describe governance practices in Dutch agricultural cooperatives (Bijman, Hendrikse, & van Oijen, 2012; Hendrikse, 2005, 2007). Given this gap in the literature, a natural extension of this line of research is to decompose ownership into distinct types of rights and to identify alternative ownership-control models of agricultural cooperatives based on the assignment of residual claims and control rights. This article begins to fill this void in the literature. Another strand of the extant research has focused on the efficiency implications of the allocation of property rights in alternative ownership structures. Among the most prominent contributions in this field is Hansmann (1996). His theory of ownership addresses the issue of why ownership of an enterprise is assigned to a particular group of patrons. Hansmann (1996) derives hypotheses dealing with the ownership costs associated with alternative ownership structures, including investor-owned, employee-owned, and farmer-owned cooperatives. However, he treats all investor-owned firms as having similar ownership rights assignments; likewise he treats all farmer-owned cooperatives as having the same ownership rights Agribusiness, Vol. 29 (1) 3–22 (2013) Published online in Wiley Online Library (wileyonlinelibrary.com/journal/agr).  C 2012 Wiley Periodicals, Inc. DOI: 10.1002/agr.21328 3 4 CHADDAD AND ILIOPOULOS assignments. He also assumes that residual claimant rights are linked to control rights, which is not always the case. In this article we address the issue of patrons with residual claimant rights in a cooperative firm transferring control rights to nonpatrons. Hansmann (1996) explicitly discusses the fact that residual control rights confer only formal control of the firm because effective control may be exercised by an individual who is not a patron (e.g., a professional manager hired as the CEO). However, he is silent about why the patrons of the firm may deliberately choose to delegate formal (and effective) control rights to nonpatron, professional managers. We address this issue by focusing on agricultural cooperatives as a unique ownership-governance structure. Understanding the rationale behind patrons’ choice to delegate both formal and real authority to nonpatron, professional managers in turn is a key to improving the efficiency of collective decision making in cooperatives. Given this backdrop, the contribution of this article is threefold. First, we shed light on alternative ownership-control models adopted by agricultural cooperatives in different regions across the world. Based on primary and secondary data sources, we describe the governance models found in cooperatives in North and South America, Northern and Southern Europe, and Oceania. In each of these models we describe the allocation of formal control rights, and in particular the allocation of decision management and decision control rights. The analysis of the formal allocation of ownership rights allows us to identify three generic governance models adopted by cooperatives in these regions: integration (in the traditional model), separation (in the extended traditional model), and delegation (in the managerial and corporate models). The distinguishing feature among these governance models is the degree of member control—that is, the extent to which member-patrons engage in decision management and decision control functions in the cooperative. Thus, our second contribution is to provide empirical evidence on the “separation of ownership and control” in agricultural cooperatives. Third, we analyze each of the governance models in terms of the associated ownership costs, including risk-bearing costs, the costs of controlling managers, and collective decision-making costs. We extend Hansmann’s (1996) theory of firm ownership and examine the ownership costs incurred by each generic governance model. In doing so, we are able to identify potential inefficiencies and understand the forces influencing the organizational efficiency and effectiveness of each cooperative model. Despite the emergence of various innovative ownership models adopted by agricultural cooperatives during recent decades, much less variance in cooperative governance models worldwide has been reported. We argue that there exist economic and organizational reasons behind this trend and propose a rationale consistent with Hansmann’s theory. The article is organized as follows. In section 2 we identify relevant concepts to the study of control rights assignment in different governance structures, then in section 3 we present evidence on the assignment of formal control rights observed in agricultural cooperatives in five regions around the world. In section 4 we introduce a conceptual framework to the analysis of control rights in alternative cooperative governance models. We present a brief sketch of Hansmann’s (1996) theory of firm ownership in the following section and extend this theory to the three generic governance models observed in most agricultural cooperatives in section 6. In section 7 we discuss the implications of our arguments, and in section 8 we conclude the paper and suggest promising avenues for future research on the governance of cooperative enterprises. 2. OWNERSHIP RIGHTS ASSIGNMENT: RELEVANT CONCEPTS Many economists would agree that the institution of ownership in the form of secure property rights is the most effective mechanism for providing economic agents with appropriate incentives to create, maintain, and improve assets. But what does “ownership” mean? The economic analysis of ownership has heretofore focused on two distinct concepts: residual returns (or claims) and residual rights of control. Agribusiness DOI 10.1002/agr CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP 5 Residual rights of control are defined as the rights to make any decision regarding the use of an asset that is not explicitly attenuated by law or assigned to other parties by contract. Residual rights of control emerge from the impossibility of crafting, implementing, and enforcing complete contracts, especially in the case of complex recurrent transactions. Because all contracts are unavoidably incomplete, it is the residual right of control over an asset that defines who is the “owner” of that asset (Grossman & Hart, 1986). Fama and Jensen (1983) argue that control rights include decision control rights (i.e., the ratification and monitoring of decisions) and decision management rights (i.e., the initiation and implementation of decisions). The assignment of decision control and decision management rights dictates what agents have formal and real (or effective) authority in organizations. Formal authority results from “an explicit or implicit contract allocating the right to decide on specified matters to a member or group of members of the organization. This formal authority, however, need not confer real authority, that is, an effective control over decisions, on its holder” (Aghion & Tirole, 1997, p. 2). According to the incomplete contract theory of the firm, the assignment of formal control rights (and hence ownership) is determined by ex ante investment incentives of contracting parties. More specifically, residual rights of control are assigned to agents making relationship specific investments whose quasirents are under risk from hold-up behavior. Hendrikse and Veerman (2001) and Hendrikse and Bijman (2002) apply this rationale in formal models predicting assignment of residual control rights to producers in a marketing cooperative and the conditions under which the marketing cooperative is more efficient than the investor-oriented firm (IOF). Economists define residual claims as the rights to the net income generated by the firm, i.e., the amount left over after all promised payments to fixed claim holders, including employees, suppliers, and debtors. Therefore, residual claimants are considered the residual risk bearers of the firm because net cash flows are uncertain and eventually negative. The “owners” of the firm are the residual claimants according to property rights and agency theory scholars (Alchian & Demsetz, 1972; Fama, 1980; Fama & Jensen, 1983). The assignment of residual claim and control rights is the basis for distinguishing between different governance structures. Table 1 summarizes the ownership rights characteristics of alternative organizational forms, including open (or publicly traded) corporations, proprietorships, financial mutual companies, and traditional cooperatives. The open corporation is characterized by unrestricted residual claims that are nonredeemable, but freely tradable in secondary equity capital markets. The horizon of residual claims is unlimited because they are rights in net cash flows for the life of the organization. In addition, residual claimants are TABLE 1. Ownership Rights Structure of Alternative Organizational Forms Open corporation Assignment of residual returns Separation of ownership from other functions Control rights Horizon of residual claims Residual claim transferability Residual claim redeemability Proprietorship Traditional cooperative Financial mutual To investors To proprietor To customers To member-patrons Yes No No No Voting rights proportional to shareholdings Unlimited Sole proprietor has control rights Customers have no control rights Nonproportional voting rights As long as proprietor As long as customer As long as patron Yes No No No No No Yes, on customer demand Yes, at Board’s discretion Note. Source: Chaddad and Cook (2004a). Agribusiness DOI 10.1002/agr 6 CHADDAD AND ILIOPOULOS not required to play any other function in the firm. The unrestricted nature of common stock residual claims enables the efficient allocation of risk and the specialization of risk bearing and decision-making functions in open corporations. In contrast to open corporations, noncorporate organizational forms usually add restrictions on residual claims that may affect asset investment and use (Table 1). We define the traditional cooperative structure as having the following property rights attributes: ownership rights are restricted to member-patrons; residual income rights are nontransferable, nonappreciable, and redeemable; and residual income is distributed to members in proportion to patronage. In addition, decision rights are exercised in a democratic way either following the one-member–one-vote rule or in proportion to patronage. This definition builds and expands on definitions provided by Hansmann (1996), Cook and Iliopoulos (1998), and Srinivasan and Phansalkar (2003). In particular, we agree that the defining attribute of any cooperative is that “members in specified exchange relationships with the cooperative are the residual claimants . . . While we use the residual claimant base as the defining characteristic [of the co-op], we will not treat any restrictions on property rights originating from cooperative principles (and often enshrined in cooperative legislation) as defining a co-op” (Srinivasan & Phansalkar, 2003, p. 372). In other words, although all cooperatives are patron-owned organizations, they adopt different ownership rights structures featuring diverse patronage, residual income, and decision rights arrangements across countries and even inside the same country. 3. ALTERNATIVE GOVERNANCE STRUCTURES IN AGRICULTURAL COOPERATIVES Cooperatives share a defining characteristic, i.e., they are patron-owned organizations, but they adopt different ownership rights structures featuring diverse patronage, residual income, and control rights arrangements. In this section we describe governance models adopted by cooperatives in different countries focusing on the allocation of decision-making functions and formal and real authority. We define generic governance models based on a number of sources. In addition to the academic literature cited, during the last 10 years we have interviewed a number of cooperative experts in several countries. The typology of European governance models is also informed by the research project Support for Farmers’ Cooperatives, which was conducted on behalf of the European Commission. The study focused, among other issues, on the governance models adopted by agricultural cooperatives in all 27 Member States of the European Union. The description of governance models in the southern cone of South America is based primarily on Brazilian agricultural cooperatives—using survey data reported in Costa, Chaddad, and Azevedo (2012) —complemented with personal interviews with managers and board members of grain, dairy, and input supply cooperatives in Argentina and Uruguay (Chaddad, 2009; Mondelli, 2012). Some of the observed variation in governance models across countries has deliberately escaped our attention due to the level of abstraction adopted in this article. 3.1. Europe 3.1.1. The Evolution of Cooperative Governance Models. Two generic models of cooperative governance are observed in Europe: the Northern European and the Southern European models (Hanisch, Rommel, & Bijman, 2012; van der Sangen & Bijman, 2012). These models are variations and extensions of a basic or traditional model (Figure 1). According to the traditional model, two decision bodies are mandatory: the general assembly (GA) and the board of directors (BoD). In some countries, a supervisory committee (SC) is also mandatory by law. In this model, decision management is performed solely by the BoD—that is, no outside agents such as professional or executive directors are delegated authority to carry out decision management functions. After being elected by the GA, BoD members allocate duties and responsibilities Agribusiness DOI 10.1002/agr CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP Figure 1 7 The Traditional Model in Europe. among themselves. It is common practice for the BoD to delegate significant real authority (all or most of decision management responsibilities) to the chairman. 3.1.1.1. General assembly. The GA of members primarily exercises ex post decision control based on an equal or proportional allocation of residual control rights. In the latter case, the allocation of voting rights is based on patronage volume or value, which can be capped. The decisions made by the GA include the election of BoD members and SC members (if the body is mandated by law or the cooperative’s bylaws), the drafting of bylaws, approval of annual reports, and decisions on major organizational changes (e.g., mergers, acquisitions, dissolution, etc.). 3.1.1.2. Board of directors. The BoD exercises ex ante decision control and decision management except on certain decisions requiring approval by the GA. Decision control rights are distributed equally to BoD members, but the chairperson may have the right to exercise veto in case of a split vote. Depending on the country, a minimum number of BoD members may be required by law (ranging from 1 to 3). The BoD is the center of both formal and real authority. Decisions are taken collectively and liabilities are borne also collectively by BoD members in the traditional governance model. 3.1.1.3. Supervisory committee. Usually SCs are not mandatory unless otherwise specified in co- operative bylaws. Traditionally, this body consists exclusively of members, but more recently some countries have allowed the appointment of nonmember (expert) participation in the SC. The main function of the SC is monitoring the BoD (ex post decision control). 3.1.2. Northern Europe. In Northern Europe, the traditional model of governance is no longer adopted by agricultural cooperatives. Instead, variations of the traditional model have been identified (Bijman et al., 2012; Brazda, 2004; Hendrikse, 2005, 2007; Kramer & Brazda, 2004; Lengsfeld, Müller, & Zieseniss, 2010; Nilsson & Ollila, 2006; Ringle, 2007). We cluster these variations into three generic models: the extended traditional, the managerial, and the corporate models. In the extended traditional model, the mandatory governance bodies are the GA and the BoD (Figure 2). A SC is mandatory in some countries when a cooperative becomes organizationally complex. Differently from the traditional model, this model introduces a nonmember agent in the cooperative governance structure. Although the BoD is in charge of strategic and Figure 2 The Extended Traditional Model in Europe. Agribusiness DOI 10.1002/agr 8 CHADDAD AND ILIOPOULOS policy decisions, all operational decisions are delegated to professional management hired by the BoD. 3.1.2.1. General assembly. In the extended traditional model, the GA of members primarily ex- ercises ex post decision control in exactly the same way as in the traditional model. In some cooperatives comprising a large and geographically dispersed membership, a system of member representation is implemented. For example, members of Arla Foods, a Swedish-Danish dairy cooperative, belong to one of sixty spatially defined wards, which in turn are grouped into seven geographic regions. Each region elects one director and 150 delegates, 140 of whom are farmer-members and 10 are employee representatives appointed by the labor unions. The 140 farmer-member delegates include 60 chairpersons from the wards, 53 ward vice chairpersons, and 27 other from the wards’ governing bodies (Nilsson & Ollila, 2006). Similar arrangements are adopted by other large cooperatives in Northern Europe (Hanisch et al., 2012). 3.1.2.2. Board of directors. The BoD exercises ex ante decision control, but unlike in the traditional model, decision management is carried out by professional management. 3.1.2.3. Supervisory committee. Usually SCs are not mandatory unless otherwise specified in co- operative bylaws. Traditionally, this body consisted exclusively of members, but more recently some countries allowed the possibility of nonmember (expert) participation. The main function is monitoring the BoD (ex post decision control). In some countries (e.g., in the Netherlands) if a cooperative builds equity capital above a certain threshold, an employee council is mandatory to ensure that the interests of employees are taken into account. Furthermore, if the cooperative employs more than a certain number of people, a board of commissioners (BoC) is also mandatory. The members of the BoC are appointed by the GA and exercise ex post decision control to ensure that the interests of all stakeholders (not just those of the patrons) are taken into consideration in decision management. BoCs are also mandatory in IOFs of the same size, but they differ from those of cooperatives, particularly in terms of the central role played by a cooperative’s GA in appointing BoC members and approving key organizational documents (Galle, 1999 in Bijman et al., 2012). 3.1.2.4. Managerial model. The managerial model eliminates one level of governance by consoli- dating the BoD and professional management. In this model, only outside professionals, who are not member-patrons of the cooperative, participate in the BoD, which is responsible for decision management functions (Figure 3). As a result, both formal and real authority reside with professional management in the managerial model. Although residual control rights for major decisions still reside with the GA, all operational and strategic decisions are delegated to professional managers. The SC (or the BoC in larger cooperatives) exercises ex post control over decisions made by the BoD. 3.1.2.5. Corporate model. The corporate model of cooperative governance consolidates the BoD and the SC or BoC (Figure 4). Both members and nonmembers (usually experts) participate in this extended BoD, but bylaws may stipulate that two-thirds of BoD members are also member-patrons of the cooperative. In this case, professional managers exercise both formal Figure 3 Agribusiness DOI 10.1002/agr The Managerial Model. CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP Figure 4 9 The Corporate Model. and real authority. Most decisions are delegated to managers and the BoD is solely responsible for decision control. 3.1.3. Southern Europe. In Southern Europe the dominant cooperative governance models are the traditional and the extended traditional as described above (Bono, 2012; Buttigieg & Zahra, 2012; Georgiou, 2012; Giagnocavo & Vargas-Vasserot, 2012; Iliopoulos, 2012; Rebelo & Caldas, 2012; Rebelo, Caldas, & Matulich, 2008; Rebelo, Caldas, & Teixeira, 2002). The traditional model of governance is adopted by the vast majority of small, local agricultural cooperatives across the Mediterranean EU countries (Figure 1). In larger cooperatives, however, the extended traditional model is implemented; professional managers are hired to perform decision management functions (i.e., the initiation and implementation of decisions). The allocation of residual claimant and control rights as well as the allocation of real and formal authority is similar to that found in the extended traditional model of Northern Europe (Figure 2). 3.2. The Southern Cone Region of South America This description of governance models in South America is based primarily on Brazilian agricultural cooperatives—using survey data reported in Costa, Chaddad, and Azevedo (2012)— complemented with personal interviews with managers and Board members of grain, dairy and input supply cooperatives in Argentina and Uruguay (Chaddad, 2009; Mondelli, 2012). Perhaps not surprisingly, farmer-owned cooperatives in the southern cone region adopt similar governance models. The Costa et al. (2012) sample includes 77 agricultural cooperatives headquartered in five different states in southern and southeastern Brazil: Minas Gerais, Paraná, Rio Grande do Sul, Santa Catarina, and São Paulo. These cooperatives have different sizes (ranging from 1 to 6,490 employees) and operate in different industries, including grain marketing, dairy processing, and farm input supply. Costa et al. (2012) observe a continuum of several distinct governance models in Brazil based on the degree of separation between risk bearing (ownership) and decision functions. Their evidence suggests the coexistence of two basic models: the traditional model and the extended traditional model, with several variations between the two depending on the extent to which member-owners engage in decision control and decision management functions. 3.2.1. Traditional Model. Principals (i.e., member-owners) delegate formal authority (both decision management and control rights) to the BoD on operational and strategic decisions, but retain decision control rights on some major decisions (mergers, acquisitions, dissolution, etc.) and ex post validation rights in the GA. In this traditional governance model, the full BoD is responsible for decision control but only a subset of BoD members is responsible for decision management (Figure 5). In general, the chairman and one or more executive directors are Agribusiness DOI 10.1002/agr 10 CHADDAD AND ILIOPOULOS Figure 5 Figure 6 The Traditional Model in South America. The Extended Traditional Model in South America. empowered with real authority. As a result, the chairman (also known as the president) plays the dual roles of the chairman of the BoD and also the CEO in this model. Costa et al. (2012) observe duality of BoD chair and CEO roles in 65% of the surveyed cooperatives. In 74% of these cooperatives with BoD chair-CEO duality, BoD members, including the chairman, are elected directly by the GA. In 26% of them, the chairman and directors with executive roles are appointed by the BoD. Following the Brazilian cooperative law, only cooperative members may be elected to the BoD, which precludes the appointment of independent, expert directors. The data from Costa et al. (2012) also show that, on average, a cooperative BoD has eight elected members with 3-year terms, with meetings on a monthly basis. Only 16% of the cooperatives in their sample have term limits on BoD directors, including the chairman. In addition to the BoD, the Brazilian cooperative law also mandates the existence of another governance body known as the supervisory board (SB) composed of at least three elected members. Nonmembers are not allowed to serve on the SB. Members of the SB may not be members of the BoD (and vice versa). The main role of the SB is to monitor the cooperative BoD and management with particular focus on internal auditing. In addition, it plays the ex post decision control function. 3.2.2. Extended Traditional Model. As in the traditional model, member-patrons delegate formal authority to the BoD, but the BoD in turn delegates real authority to the CEO. The CEO may be a hired professional or a member of the cooperative. In this model, there is separation of decision control (retained by the BoD) and decision management (by the CEO) and there is no BoD Chair-CEO duality (Figure 6). Costa et al. (2012) find that 27% of the surveyed cooperatives have partial separation and only 8% have complete separation of decision control from decision management functions. Following Brazilian law, this model also includes a SB. 3.3. North America This description of governance models in North America is based primarily on U.S. agricultural cooperatives—using survey data collected by Burress, Livingston, and Cook (2011). The dominant governance model followed by U.S. agricultural cooperatives is the extended traditional model with a clear separation of decision control and decision management functions. In this Agribusiness DOI 10.1002/agr CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP 11 governance model, the BoD exercises decision control, whereas decision management is the responsibility of a professional management team lead by the CEO. We are not aware of any farmer-owned cooperative in the United States with duality of CEO and chairman roles, but in a few exceptional cases, the CEO may also have a seat on the BoD. The data reported in Burress et al. (2011) was collected from a mail survey of BoD chairmen. The sample comprised the top 1,000 cooperatives in the U.S. Department of Agriculture Cooperative Statistics database. Completed questionnaires were received from 460 survey respondents, including marketing, input supply, service, multipurpose, and new generation cooperatives. The average size of these cooperatives was $252 million in turnover, ranging from $528,000 to $26 billion. The average number of fulltime employees was 195, ranging from 0 to 9,739. The average size of the cooperative BoD is nine, ranging from five to 51 directors. A striking characteristic of these boards is that only 4.4% of the U.S. cooperatives have at least one outside, nonmember director and only 2% grant voting rights to outside directors. In other words, decision control in U.S. cooperative boards is exercised primarily by elected members with little help from outside directors. 3.4. Oceania The governance model adopted in the majority of agricultural cooperatives in Australia and New Zealand is very similar to the extended traditional model described for their U.S. counterparts. There is a clear separation between decision control exercised by the BoD and decision management by the CEO and the senior management team. One notable difference is that in most cooperatives in Oceania the BoD has a mixed structure with both farmer-elected BoD members and outside, independent directors (Table 2). Even though the selected cooperatives in Table 2 are not representative of the population, they include cooperatives of different sizes, with operations in different industries and with different residual claim structures. Table 2 suggests that cooperatives in Oceania tend to adopt a BoD structure with participation of outside directors with complementary skills to farmer-directors. Plunkett, Chaddad, and Cook (2010) also find the presence of independent board directors in dual-structured irrigation cooperatives in Australia. State Co-operatives Acts in Australia determine that the majority of BoD directors must be active members, but the BoD may include employees of the cooperative and nonmembers if permitted by the cooperative bylaws. Most cooperatives in Australia adopt the one-member–one-vote rule; however, cooperatives in New Zealand tend to favor proportional voting on the basis of patronage volume. It is worth noting that New Zealand cooperatives are formed under special provisions of the Company’s Act, which is the basis for proportional voting. Similarly, Murray Goulburn in Australia is formed under federal company legislation and follows limited proportional voting. Cooperatives formed under State cooperative TABLE 2. Composition of the Board of Directors (BoD) in Selected Cooperatives in Oceania Cooperative CBH Murray Goulburn Namoi Cotton Norco Fonterra LIC Ravensdown Tatua Westland a Country BoD Size Farmer-elected BoD members Australia Australia Australia Australia New Zealand New Zealand New Zealand New Zealand New Zealand 12 11 7 7 13 10 14 8 10 9 9 4 6 9 7 12 7 8 Appointed, outside directors 3 a 2 3 1 4 3 2 1 2 One independent director and the managing director. Agribusiness DOI 10.1002/agr 12 CHADDAD AND ILIOPOULOS legislation, such as CBH, Norco, and Namoi, are required to adhere to the one-member–onevote principle. The majority of farmer-owned cooperatives in Oceania adopt a one-tier governance structure with the BoD being the sole governance body in addition to the GA. Fonterra and LIC are outliers as they follow a two-tiered governance model with a BoD and a Shareholders’ Council (SC). Following common practice in the continent, Fonterra’s BoD is comprised of both farmers elected by members and appointed directors with special skills. The nine farmer directors are elected at large–that is, they do not represent any region–with proportional voting as Fonterra adopts a proportional ownership model that requires members to buy one share for each kilogram of milk solids delivered. The SC in turn is comprised of 35 elected farmers following the one-member–one-vote rule. Each elected councilor represents one specific geographic region (i.e., a ward) in New Zealand. In addition to member representation, the role of the SC includes monitoring the performance of the cooperative, appointing the valuer of the fair value share and the milk commissioner, setting the rules for and managing the election of BoD members, and developing future leaders. Also, any proposed constitutional changes initiated by the BoD must first pass scrutiny by the SC before going to the membership for a vote. 4. MEMBER CONTROL ALONG THE GOVERNANCE STRUCTURE CONTINUUM An analysis of the governance models described above—based on the allocation of decision management and decision control rights—enables us to examine the “separation of ownership and control” in agricultural cooperatives. Our synthesis suggests that most farmer-owned cooperatives follow one of three broad types of governance models along a “member control” continuum: quasi-integration, separation, and delegation (Figure 7). The polar form on this continuum is integration, which occurs when principals take responsibility for both decision control and decision management, and therefore retain formal Figure 7 Agribusiness DOI 10.1002/agr Governance Structure Continuum. CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP 13 and real authority. In other words, there is no “separation of ownership from control” (Berle & Means, 1932) or no “separation between risk bearing and decision management” (Fama & Jensen, 1983). Integration occurs in sole proprietorships (by default) and also in general partnerships when the number of partners is sufficiently small for all of them to be involved in decision management. It is only in this polar case that principals—i.e., member-patrons in cooperatives—would retain 100% control. However, as the number of principals increases, integration is not workable as it would lead to slow decision making and high collective decision making costs (Hansmann, 1996). This is the case in most complex organizations with dispersed ownership such as cooperatives. A workable model for many small cooperatives is quasi-integration, when principals delegate formal and real authority to the BoD. The BoD may in turn delegate decision management responsibilities to a subset of executive directors or enshrine a member (i.e., the president) with a great deal of power as observed in many South American cooperatives. The traditional governance model that is dominant in South America and Mediterranean countries would be classified as quasi-integration. Quasi-integration suggests that most members delegate decision rights and authority to a small group of member-patrons and therefore lose some degree of control. Quasi integration lies to the right of integration in the governance structure continuum. The next governance structure along the continuum is separation. In this model, principals delegate formal authority to the BoD, which focuses on decision control but delegates decision management rights to professional managers. In other words, there is a clear separation between risk bearing and decision management and also separation between formal and real authority. Separation suggests, therefore, that principals are delegating more decision making power with consequent reduction in the level of member control. The extended traditional model that is dominant in North America and Oceania, and is also very common in Northern Europe, would be an example of separation. A more muted level of member control is delegation, i.e., the principals give up substantial control (including formal and real authority) to agents. Principals retain ex post control rights, which may be thought of as a type of contingent control claim in case the agent’s decisions lead to poor performance. As explained by Hendrikse (2005, p. 389), “members delegate formal rights to professional management as long as everything works well, while these rights go back to the members during bad times.” The corporate and managerial models adopted by some Dutch agricultural cooperatives (Bijman et al., 2012) are examples of this delegation governance structure. Lastly, cooperative members give up control of downstream or upstream assets and engage in arms-length market contracting with an IOF when they decide to demutualize (Chaddad & Cook, 2004b). In doing so, they avoid the ownership costs of risk bearing, control and collective decision making but incur market contracting costs in input supply procurement and marketing their farm output (Hansmann, 1996). Why would principals give up control? In other words, why are all cooperatives not 100% member controlled? One possible explanation is related to the high transaction costs incurred by members in controlling their cooperative (Staatz, 1987). Along the same lines, another likely explanation is that the perceived benefits of delegating control are higher than the corresponding costs for member-owners, including the aforementioned transaction costs. As is well established in the literature, agency costs arise when decision agents do not bear the wealth effects of their decisions (Jensen & Meckling, 1976). The benefits of separating risk-bearing from decision-management functions include gains from specialization, the hiring of professional managers, and more effective use of specific knowledge in complex organizations (Fama & Jensen, 1983). In addition, delegating authority increases the agent’s initiative and facilitates the agent’s participation in the contractual relationship with principals (Aghion & Tirole, 1997). This trade-off along the governance structure continuum suggests the relevance of exploring the determinants of governance choice—that is, what are the factors that explain what governance model is chosen? We attempt to inform this issue in the remainder of this article. Agribusiness DOI 10.1002/agr 14 CHADDAD AND ILIOPOULOS 5. HANSMANN’S THEORY OF FIRM OWNERSHIP Hansmann’s (1996) theory of enterprise ownership posits that over the long run, costminimizing forms of organization will dominate most industries. Costs include both monetary values, but also interests and values that might be affected by transactions between a firm and its patrons. These costs are grouped into two categories: the costs of market contracting and the costs of ownership. If there are N classes of patrons who transact with the firm, efficiency dictates that ownership should be assigned to that class that minimizes the sum: CO j +  CiKj , i= j K where C O j is the cost of ownership for the group of patrons in class i and Ci j is the cost of market contracting for the group of patrons in class i when class j owns the firm. The costs of market contracting arise as a result of various forms of market failures. By assigning ownership to the affected parties, these costs can be reduced significantly. Eight types of contracting costs are identified: (a) simple market power, (b) ex post market power, (c) the risks of long term contracting, (d) asymmetric information, (e) strategic bargaining, (f) communication of patron preferences, (g) compromising among diverse patron interests, and (h) alienation. Market-contracting costs have a tremendous influence on the decision of a group of farmers to form an agricultural cooperative to ameliorate perceived market failures. The minimization of these costs has provided the rationale for the formation of most traditional agricultural cooperatives across the globe in the early 1900s. However, the various market contracting costs do not seem to have a significant impact on the choice of a cooperative’s governance model. The costs of ownership, on the other hand, crucially affect the governance model a cooperative chooses to adopt. Such costs are conveniently grouped into the costs of controlling managers, the costs of collective decision making, and the costs of risk bearing. In this article we focus the analysis of governance models on the costs of controlling managers and the costs of collective decision making, which are more directly related to the allocation of control rights, and thus leave the implications to risk-bearing costs to future research. Controlling managers is associated with significant costs incurred by the patrons of a firm. These include the costs of monitoring the management and costs resulting from managerial opportunism. In monitoring the management efficiently, patron-owners need to inform themselves about the operations of the firm, communicate among themselves for the purpose of exchanging information and making decisions, and bring their decisions to bear on the firm’s management. The resulting monitoring costs for a given class of patrons will generally be inversely proportional to the importance, frequency, and duration of the patrons’ transactions with the firm. The costs of monitoring will also depend on the ease of organizing the patronowners for collective action that, in turn, depends on factors such as the patrons’ physical proximity to one another and to the firm. As the number of owners grows, the share of each individual owner of the potential gains from effective monitoring decreases with a consequent reduction in incentives to monitor the management. In explaining the contrasting reality of firms with a very large group of owners, it is asserted that either or both of two things must happen: (a) the costs of market contracting would be much higher under any alternative ownership assignment, (b) the costs of managerial opportunism are modest even though the firm’s owners cannot actively supervise the managers. Indeed, in analyzing the costs arising from managerial opportunism, the theory of enterprise ownership argues that there are important limits to such costs even in firms whose nominal owners are in a poor position to monitor the firm’s management at all. In addition to the agency costs of controlling managers, firms also incur collective decisionmaking costs that result from heterogeneity of interests among the firm’s owners. Interest heterogeneity emerges due to differences in the way in which patron-owners transact with the Agribusiness DOI 10.1002/agr CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP 15 firm, or as a result of differences in personal circumstances. Collective decision-making costs arise with the adoption of costly processes to deal with owners’ differences and also as a result of inefficient decisions, whose outcomes fail to maximize the aggregate welfare of the owners as a group. Poor decisions arise because of inefficient voting systems that fail to select outcomes preferred by the average owner or result in control falling into the hands of an unrepresentative minority whose decisions inefficiently exploit the majority in its favor. The latter is more likely to happen when some patrons are better situated to participate effectively in collective decision making than others. The process of collective decision making also may impose significant costs on the firm’s owners. An owner may bear significant costs to obtain knowledge about the firm and the preferences of other owners, and also to attend meetings and related activities necessary to reach and implement effective decisions. Another source of costs are voting cycles among alternative options; the incidence of voting cycles increases as preferences among the electorate become increasingly heterogeneous. Further, voting cycles may result in extraordinary power being seized by a subset of owners who do not care about the efficiency of decision outcomes. Finally, in case owners behave strategically, additional costs may result from efforts to hide or discover information or to make or break coalitions. The costs of costly decision-making processes may be reduced significantly by implementing various methods. The delegation of authority to committees is proposed as a means of reducing the costs of participation, inhibiting vote cycling, and facilitating logrolling that would mitigate the median voter problem. Even in the case of highly heterogeneous member interests collective decision making costs may be kept low if a simple and salient criterion for balancing those interests is adopted. 6. OWNERSHIP COSTS ACROSS COOPERATIVE GOVERNANCE MODELS 6.1. Applying Hansmann’s Theory Hansmann (1996) applies his theoretical framework to the study of various forms of business ownership in the United States, including farmer-owned cooperatives. Both in his theoretical and empirical analyses, it is assumed that all cooperatives are characterized by the same assignment of ownership rights. As a result, the theory oversees the issue of how the costs of ownership might be increased or reduced by adopting nontraditional models. Neither does the theory address the issue of why a firm’s nominal owners may deliberately choose to delegate both formal and real authority to professional managers. In this section, we expand on both of these subjects. Based on our experience and the extant academic literature we provide a preliminary assessment of the various types of ownership costs identified by Hansmann (1996) for each of the governance models described in section 2. We assume that a cooperative’s adoption of the extended traditional, managerial, or corporate model is associated with, though not necessarily caused by, increased organizational complexity. Organizational complexity arises primarily due to changes in cooperatives’ strategies and structures that either result in highly heterogeneous member interests or alter significantly the way in which member-patrons transact with their cooperative. 6.2. The Costs of Controlling Managers 6.2.1. Costs of Monitoring the Management. Three factors influence monitoring costs: (a) the degree of heterogeneity of members’ interests; (b) the importance, frequency, and duration of members’ transactions with the cooperative; and (c) members’ capability to comprehend complex issues under the assumption of bounded rationality.1 The monitoring costs associated with the traditional governance model are presumably very low because it is exclusively board 1 We thank the editors Jerker Nilsson and Petri Ollila for bringing this last factor to our attention. Agribusiness DOI 10.1002/agr 16 CHADDAD AND ILIOPOULOS directors elected by members that assume all managerial duties. Given that this governance model is primarily adopted by local, less organizationally complex cooperatives, informing members about the operations of their cooperative is relatively costless. Communication among members for the purpose of exchanging information and making decisions, as well as bringing members’ decisions to bear on the cooperative’s management, also do not pose significant challenges. Thus the associated costs are generally low. As members’ interests diverge more and more, organizing and informing member-patrons become increasingly difficult. Also, the costs of monitoring will generally be inversely proportional to the importance, frequency, and duration of member-patrons’ transactions with the cooperative (Hansmann, 1996). Further, as the number of members increases and as these members are scattered over a larger geographic area, individual members will tend to free ride on the monitoring efforts of other members (Hansmann, 1996; Nilsson & Svendsen, 2011; Olson, 1965). Further, as the cooperative increasingly faces more complex decisions, members may find it difficult to understand the various aspects of the issues involved. As a result, monitoring efficiency will fall while the associated costs will rise. Most likely, the aforementioned factors become progressively more relevant in cooperatives adopting the extended traditional, or even more so, the managerial or corporate governance model. 6.2.2. Costs of Managerial Opportunism. The costs of managerial opportunism can be ex- tremely low in the traditional model, if and only if cooperative leaders are highly efficient and nonopportunistic (Iliopoulos & Valentinov, 2012; Nilsson, 2001). This is in line with the very low agency costs that Hansmann (1996) observes in traditional agricultural cooperatives in the United States. He argues that there are important limits to the costs of managerial opportunism even when member-owners are in a poor position to do any active monitoring of the cooperative’s management at all. Accepting this argument implies that these costs are low, irrespective of the governance model adopted. Yet, as new governance models that increasingly delegate real authority to professional managers emerged during recent decades, the costs of managerial opportunism are expected to rise. Hansmann’s (1996) claim that the composition of cooperatives’ boards of directors on which only members serve reflects a high degree of member control is challenged by the adoption of the aforementioned governance models where professional managers dominate the boards. Similarly challenged is Hansmann’s (1996) argument about the role of federated structures in enabling effective control of managers by representatives of local, first-tier cooperatives. The gradual abandonment of the federated model in favor of centralized structures in many countries suggests that either centralized cooperatives provide an equally efficient means to monitoring managers or other considerations might have been more important in choosing among models. The demise of the once mighty federated Dutch cooperative, Royal Cebeco, in the early 2000s serves as an illustrative example (Bijman, 2005). Representatives of Dutch cooperatives view the loss of member control as a consequence of adopting the managerial or the corporate model; one that poses a potential threat to the future success of their organization (Bijman et al., 2012). In a different institutional setting, in new generation cooperatives, which mainly adopt the extended traditional governance model, the costs of managerial opportunism seem to be partially mitigated by the monitoring effects of the secondary market for delivery rights. 6.3. The Costs of Collective Decision Making The traditional model of governance is associated with low costs of collective decision making due to members’ interests being, in general, homogeneous. Yet, as cooperatives progress along their life cycle, members’ interests become increasingly heterogeneous (Cook, 1995). Consequently, collective decision making may become onerous and thus impose significant costs on cooperatives. Slow decision making, costly decisions, and costly processes during this phase are what farmers commonly refer to as “cooperative baggage.” The adoption of various Agribusiness DOI 10.1002/agr CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP TABLE 3. 17 Control Costs in Alternative Governance Models in Cooperatives Governance model Costs associated with exercise of control Traditional model Costs of monitoring the management Very low Members informing themselves about cooperative operations Low to medium Members communicating to exchange information and make decisions Members bringing their decisions to bear on the cooperative’s management Very low to low Costs of managerial opportunism Zero, if cooperative leaders are efficient and not opportunistic Costs of collective decision making Costs of decisions Low Costs of process Low Costs of resolving Low conflicts Extended traditional model Low to medium Managerial and corporate models Medium to high, depending on importance, frequency, and duration of members’ transactions with the co-op. Low to medium, depending on (a) the composition and effectiveness of the BoD, and (b) the adoption of clearly defined performance measures Medium to high, depending on degree of heterogeneity of members’ interests Medium to high, depending on (a) degree of heterogeneity of members’ interests; and (b) importance, frequency, and duration of members’ transactions with the cooperative Medium to high, depending on importance, frequency, and duration of members’ transactions with the cooperative Medium to high, depending on (a) the composition and effectiveness of the BoD, and (b) the adoption of clearly defined performance measures Low to medium Low to medium Low to medium, depending on degree of heterogeneity of members’ interests. Very low Very low Very low, depending on degree of heterogeneity of members’ interests Medium to high governance schemes and incentive mechanisms by cooperatives with heterogeneous memberships may represent attempts to lessen the costs of collective decision making. As argued by Bijman et al. (2012), the adoption of the managerial or corporate models of governance by several Dutch agricultural cooperatives might have been necessitated by the need to implement efficient decision making routines when too much member involvement would jeopardize organizational efficiency. Table 3 summarizes the above arguments and provides an assessment of the hypothesized ownership costs incurred under each generic governance model. 7. DISCUSSION Table 3 suggests that as organizational complexity increases agricultural cooperatives tend to move from the traditional to the extended traditional to the managerial and corporate Agribusiness DOI 10.1002/agr 18 CHADDAD AND ILIOPOULOS models of governance. Switching to alternative governance models also implies a tradeoff. What cooperatives lose, in terms of additional monitoring costs incurred, they gain in terms of lower collective decision-making costs. However, whether they actually realize additional monitoring costs and lower collective decision making costs ultimately depends on a number of factors, including (a) the degree of homogeneity of members’ transactions with the cooperative, (b) the degree of homogeneity of members’ preferences, (c) the costs associated with the adoption of efficient rules and routines to balance potentially conflicting member interests (e.g., rules for allocation of overhead costs), (d) the composition and effectiveness of the board of directors, and (e) the implementation of clearly defined performance measures. In considering the minimization of ownership costs, the move from the traditional to the extended traditional, managerial, and corporate models suggests that the costs of collective decision making are far more important than the sum of the costs of monitoring the management and those incurred due to managerial opportunism. In cooperatives with highly heterogeneous member interests, the efficiency gains resulting from the adoption of a nontraditional governance model seem to depend on the costs associated with the implementation of processes and routines that resolve intracooperative conflicts. Thus the emergence of numerous, innovative ownership arrangements during the last 20 years might also have had the indirect effect of ameliorating the negative impact of divergent member interests (Chaddad & Cook, 2004a). This may also explain the lack of significant variation in governance models observed by Bijman et al. (2012) and further analyzed in this article. In other words, innovations in cooperative ownership may also have adequately addressed governance challenges by homogenizing members’ interests. The organizational efficiency and long-term viability of each of the three governance models are facilitated by a high degree of homogeneity in members’ interests and efficient, nonopportunistic leaders. In addition to these factors, ceteris paribus, the traditional governance model has a higher chance of success when membership is comprised by a relatively low number of members who are geographically concentrated. Cooperatives adopting the extended traditional model need also carefully design and implement operational and constitutional rules that minimize potential conflicts among members with increasingly divergent interests. The adoption of separate patronage, capital, governance, and risk pools is an example of such a mechanism (Cook & Iliopoulos, 1998). The ability of cooperatives adopting the managerial or corporate model to implement highly efficient monitoring mechanisms seems to be a prerequisite for survival. Such cooperatives may also need to implement microgovernance mechanisms that will ensure that members’ interests are taken into account. As agricultural cooperatives are increasingly challenged by divergent member interests, the last policy may become even more important. Along these lines, the composition and effectiveness of the board of directors is yet another factor that plays a crucial role in minimizing the costs of ownership incurred by agricultural cooperatives adopting the managerial or corporate governance models. Although the empirical literature is far from providing us with a definite list of board characteristics associated with increased firm performance, many studies seem to agree on the importance of these characteristics in corporations (Hermalin & Weisbach, 2003). For example, choosing who and how many individuals will serve on a particular cooperative’s board has significant efficiency implications and may alter monitoring costs accordingly (Bond, 2009). A closely related issue is the design and implementation of clearly defined performance measures that enable the board and members to monitor the management. In this way, board members avoid confusion over what metric to use when evaluating managers (Richards, Klein, & Wallburger, 1998) and are enabled to design compensation schemes aligning the objectives of the management with those of members (Staatz, 1983). Further, by increasing transparency, clearly defined performance measures allow members, particularly in cooperatives with a large, geographically dispersed membership, to monitor management resourcefully. This is even more important in cooperatives adopting the managerial or corporate models with delegation of formal and real authority to professional managers. Agribusiness DOI 10.1002/agr CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP 19 Lastly, the choice of governance model is embedded into a cooperative’s institutional environment. In other words, the above discussion assumes a constant, exogenously determined institutional environment. However, the influence of the formal rules of the game on the choice of governance model seems to be crucial. For example, Bijman et al. (2012) suggest that the flexibility afforded by Dutch laws may explain cooperative’s experimentation with alternative governance models. Further, the predominant cooperative structure in the Netherlands, where a wholly owned IOF is the business branch of the cooperative association, is closer to adopting the managerial or corporate governance model than are the traditional, single-entity cooperatives found in South America, Southern Europe, and the United States. Informal institutions, such as customs, norms, routines, and political processes also influence the choice of governance model significantly (Zenger, Lazzarini, & Poppo, 2002). 8. CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH Agricultural cooperatives experiment with innovative ownership and governance structures to increase their efficiency and survive in increasingly competitive environments. Among the recently observed changes in cooperative governance is the Dutch case where several agricultural cooperatives have delegated formal and real authority to nonpatron, professional managers. Does this case represent an outlier or will cooperatives around the globe start to follow their Dutch counterparts? Here we have analyzed three generic governance models currently adopted by cooperatives in various countries in terms of the resulting ownership costs identified in Hansmann’s (1996) theory of enterprise ownership. Based on our analysis, the following conclusions can be drawn: 1. As organizational complexity increases, cooperatives modify their traditional ownership structure to attract risk capital and minimize potential conflicts among members with increasingly diverging interests. If cooperatives are successful in pursuing this strategy, then members might oppose the delegation of formal and real authority to managers. If, however, organizational complexity is addressed through other means (e.g., a cooperative-owned investor-oriented firm as is common practice in the Netherlands), then member-patrons may accept that formal and real control of this separate entity should be entrusted to professional managers. 2. As cooperatives transition from the traditional to the extended traditional, the managerial or corporate governance model, ceteris paribus, the costs of controlling managers increase while the costs of collective decision making are potentially reduced. This is a consequence of increasing levels of control being delegated to nonpatron managers. 3. Although in traditional agricultural cooperatives the costs of managerial opportunism are relatively low, the same claim cannot be made with certainty for the respective costs incurred by cooperatives adopting the managerial or corporate governance models. 4. In many cases, the increased organizational complexity caused by or resulting in a higher degree of heterogeneity in members’ interests has rendered the federated cooperative structure obsolete with a consequent shift to increasingly centralized structures. 5. There is a constant interplay between the incentives created by changes in the ownership structure, the degree of membership heterogeneity, and the governance model adopted by a cooperative. One cannot tinker with one of these three without causing changes in the other two. 6. Above a certain level of organizational complexity, not switching to one of the innovative ownership models described by Chaddad and Cook (2004a) may force cooperative members to delegate more real authority to nonpatron managers. In particular, cooperatives that have adopted one of these models (e.g., new generation cooperatives) are less in need of tinkering Agribusiness DOI 10.1002/agr 20 CHADDAD AND ILIOPOULOS with their governance model. In turn, this may explain the lack of significant variance in the governance models adopted by cooperatives worldwide. 7. In cooperatives with highly heterogeneous member interests, the efficiency gains resulting from the adoption of nontraditional governance models depend on the costs associated with the implementation of rules, processes, and routines that resolve intracooperative conflicts. 8. The adoption of the managerial and corporate governance models by several Dutch cooperatives in recent years is an experiment that will ultimately test Hansmann’s (1996) hypothesis that managerial opportunism is insignificant even when a firm’s nominal owners do not exercise effective monitoring of the management. These conclusions can serve as hypotheses in future empirical research investigating the governance of farmer-owned cooperatives. The governance of agricultural cooperatives has heretofore received relatively little attention from theoretical scholars. Yet, it is theoretically informed empirical work that will help us shed new light on the microinstitutions of governance that make a significant difference. Both case study approaches and statistical models will be useful in identifying variations of the three generic governance models, but also in understanding how the interplay among a particular ownership structure, the degree of membership heterogeneity, and the governance model adopted by a cooperative affects organizational efficiency. REFERENCES Aghion, P., & Tirole, J. (1997). Formal authority in organizations. Journal of Political Economy, 105, 1–29. Alchian, A.A., & Demsetz, H. (1972). Production, information costs, and economic organization. American Economic Review, 62, 777–795. Berle, A., & Means, G. (1932). The modern corporation and private property. New York: Macmillan. Bijman, J. (2005). Network and hierarchy in Dutch co-operatives: A critical analysis. The International Journal of Co-operative Management, 2, 16–24. Bijman, J., Hendrikse, G., & van Oijen, A. (2013). Accommodating two worlds in one organization: Changing board models in agricultural cooperatives. Managerial and Decision Economics, vol. 34, forthcoming. Bond, J.K. (2009). Cooperative financial performance and board of director characteristics: A quantitative investigation. Journal of Cooperatives, 22, 22–44. Bono, P. (2012). Support for farmers’ cooperatives. Unpublished Country Report Italy, Wageningen UR, The Netherlands. Brazda, J. (2004). Perspectives of agricultural co-operatives in Austria. In J. Brazda (Ed.), Agricultural co-operatives are facing a challenge (pp. 5–28). Vienna, Austria: University of Vienna, Department of Business Studies. Burress, M., Livingston, K., & Cook, M.L. (2011, Fall). Cooperative boards: A descriptive summary of survey findings regarding demographics, director development and member engagement. The Cooperative Accountant, pp. 20–30. Buttigieg, G., & Zahra, E. (2012). Support for farmers’ cooperatives. Unpublished Country Report Malta, Wageningen UR, The Netherlands. Chaddad, F.R. (2009). The Uruguayan dairy sector in an international context: Organization and sector strategy. Unpublished Research Report, Food and Agriculture Organization (FAO) of the United Nations, Rome, Italy. Chaddad, F.R., & Cook, M.L. (2004a). Understanding new cooperative models: An ownership-control rights typology. Review of Agricultural Economics, 26, 348–360. Chaddad, F.R., & Cook, M.L. (2004b). The economics of organization structure changes: A U.S. perspective on demutualization. Annals of Public and Cooperative Economics, 75, 575–594. Cook, M.L. (1995). The future of U.S. agricultural cooperatives: A neo-institutional approach. American Journal Agricultural Economics, 77, 1153–1159. Cook, M.L., & Iliopoulos, C. (1998). Solutions to property rights constraints in producer owned and controlled organizations: Prerequisite for agri-chain leadership? In G.W. Ziggers, J.H. Trienekens, & P.J. Zuurbier (Eds.), Third International Conference on Chain Management in Agribusiness and the Food Industry (pp. 541–553). The Netherlands: Wageningen Agricultural University. Cook, M.L., & Iliopoulos, C. (2000). Ill-defined property rights in collective action: The case of agricultural cooperatives. In C. Menard (Ed.), Institutions, contracts, and organizations: Perspectives from new institutional economics (pp. 335–348). London: Edward Elgar. Costa, D.M., Chaddad, F.R., & Azevedo, P.F. (2012). Separation of ownership and decision management in Brazilian agricultural cooperatives. Revista de Economia e Sociologia Rural, 50, 285–300. Fama, E.F. (1980). Agency problems and the theory of the firm. Journal of Political Economy, 88, 288–307. Agribusiness DOI 10.1002/agr CONTROL RIGHTS, GOVERNANCE, AND THE COSTS OF OWNERSHIP 21 Fama, E.F., & Jensen, M.C. (1983). Separation of ownership and control. Journal of Law and Economics, 26, 301– 326. Georgiou, E.K. (2012). Support for farmers’ cooperatives. Unpublished country report Cyprus, Wageningen UR, The Netherlands. Giagnocavo, C., & Vargas-Vasserot, C. (2012). Support for farmers’ cooperatives. Unpublished country report Spain. Wageningen UR, The Netherlands. Grossman, S.J., & Hart, O.D. (1986). The costs and benefits of ownership: A theory of vertical and lateral integration. Journal of Political Economy, 94, 691–719. Hanisch, M., Rommel, J., & Bijman, J. (2012, March). Governance and performance-analyzing patterns of internal governance in agricultural cooperatives across the EU-27. Paper presented at the International Conference Cooperative Responses to Global Challenges, Berlin, Germany. Hansmann, H. (1996). The ownership of enterprise. Cambridge: The Belknap Press of Harvard University Press. Hendrikse, G. (2005). Contingent control rights in agricultural cooperatives. In T. Theurl & E.C. Meyer (Eds.), Strategies for Cooperation (pp. 385–393). Maastricht: Shaker Verlag. Hendrikse, G. (2007). Two vignettes regarding boards in cooperatives versus corporations. In K. Karantininis & J. Nilsson (Eds.), Vertical markets and cooperative hierarchies (pp. 137–150). Dordrecht: Springer. Hendrikse, G., & Veerman, C.P. (2001). Marketing co-operatives: An incomplete contracting perspective. Journal of Agricultural Economics, 52, 53–64. Hendrikse, G., & Bijman, J. (2002). Ownership structure in agrifood chains: The marketing cooperative. American Journal of Agricultural Economics, 84, 104–119. Hermalin, B.E., & Weisbach, M.S. (2003). Boards of directors as an endogenously determined institution: A survey of the economic literature. Economic Policy Review 9, 7–26. Holland, S.J., & King, R.P. (2004). Trading mechanisms for new generation cooperative stock: The architecture of organizational formation and demise. American Journal of Agricultural Economics, 86, 1262–1268. Iliopoulos, C. (2012). Support for farmers’ cooperatives. Unpublished country report Greece, Wageningen UR, The Netherlands. Iliopoulos, C., & Valentinov, V. (2012). Opportunism in agricultural cooperatives in Greece. Outlook on Agriculture, 41, 15–19. Jensen, M.C., & Meckling, W.H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–360. Kramer, J.W., & Brazda, J. (2004). Corporate governance in German rural co-operatives from a property rights point of view. In J. Brazda (Ed.), Agricultural co-operatives are facing a challenge (pp. 29–61). Vienna, Austria: University of Vienna, Department of Business Studies. Lengsfeld, S., Müller, D., & Zieseniss, R. (2010). Group and shareholding structures of agricultural, input supply and service cooperatives in Germany 60, 290–304. Mondelli, M.P. (2012). Analisis de las innovaciones en las cooperativas de granos del Uruguay. Unpublished research report, Centro de Investigaciones Económicas (CINVE), Montevideo, Uruguay. Nilsson, J. (1999). Cooperative organisational models as reflections of the business environments. The Finnish Journal of Business Economics, 4, 449–470. Nilsson, J. (2001). Organisational principles for cooperative firms. Scandinavian Journal of Management, 17, 329– 356. Nilsson, J., & Ollila, P. (2006). The dairy co-operatives in the Nordic countries. Report to the National Dairy Development Board of India. Uppsala, Sweden: Swedish University of Agricultural Sciences, Department of Economics. Nilsson, J., & Svendsen, G.T. (2011). Free riding or trust? Why members (do not) monitor their cooperatives. Journal of Rural Cooperation, 39, 131–150. Olson, M. Jr. (1965). The logic of collective action. Cambridge, MA: Harvard University Press. Plunkett, B., Chaddad, F.R., & Cook, M.L. (2010). Ownership structure and incentives to invest: Dual structured irrigation cooperatives in Australia. Journal of Institutional Economics, 6, 261–280. Rebelo, J., & Caldas, J. (2012). Support for farmers’ cooperatives. Unpublished country report Portugal. Wageningen UR, The Netherlands. Rebelo, J., Caldas, J., & Matulich, S. (2008). Manager power, member behaviour and capital structure: Portuguese Douro wine cooperatives. Agricultural Economics Review, 9, 5–13. Rebelo, J., Caldas, J., & Teixeira, M. (2002). Economic role, property rights, labour skills and technology in Portuguese wine cooperatives. Annals of Public and Cooperative Economics, 73, 111–133. Richards, T.J., Klein, K.K., & Wallburger, A. (1998). Principal-agent relationships in agricultural cooperatives: An empirical analysis from rural Alberta. Journal of Cooperatives, 13, 21–34. Ringle, G. (2007). Cooperative principles in the field of tension between tradition and modernity. (Working paper Heft 01/2007). Germany: Hochschule Wismar, University of Technology Business and Design. Srinivasan, R., & Phansalkar, S.J. (2003). Residual claims in co-operatives: Design issues. Annals of Public and Cooperative Economics, 74, 365–395. Staatz, J.M. (1983). The cooperative as a coalition: A game-theoretic approach. American Journal of Agricultural Economics, 65, 1084–1089. Agribusiness DOI 10.1002/agr 22 CHADDAD AND ILIOPOULOS Staatz, J.M. (1987). Farmers’ incentives to take collective action via cooperatives: A transaction-cost approach. In J.S. Royer (Ed.), Cooperative theory: New approaches (pp. 87–107) (Agricultural Cooperative Service Report 18). Washington, DC: U.S. Department of Agriculture. Van der Sangen, G., & Bijman, J. (2012). Internal governance of cooperatives. Unpublished report, Wageningen UR, The Netherlands. Zenger, T.R., Lazzarini, S.G., & Poppo, L. (2002). Informal and formal organization in new institutional economics. In P. Ingram & B.S. Silverman (Eds.), The new institutionalism in strategic management: Advances in strategic management (Vol. 19, pp. 277–305). Bingley, UK: Emerald Group Publishing Limited. Fabio Chaddad is an assistant professor of agricultural economics at the Division of Applied Social Sciences, University of Missouri. He has a PhD in Agricultural Economics from the University of Missouri (2011), a master’s degree in Agribusiness Management from the University of Sao Paulo, Brazil (1996) and a bachelor’s degree in Agriculture also from the University of Sao Paulo (1992). He currently teaches strategic management, corporate governance, and organizational economics at the undergraduate and graduate levels. Fabio’s research activities focus on the economics and management of user-owned and controlled organizations and interfirm collaborative arrangements in the global agrifood system. Constantine Iliopoulos is an expert on the organization and economics of food and agribusiness supply chains and cooperatives. Currently, he is a researcher with the Agricultural Economics and Policy Research Institute of the National Agricultural Research Foundation, and an adjunct professor at the Department of Agricultural Economics and Rural Development of the Agricultural University of Athens, Greece. Dr. Iliopoulos received his BA and MS in Agricultural Economics from the Agricultural University of Athens, Greece (1992), and his PhD in Agribusiness Economics from the University of Missouri at Columbia (1998). His research program has focused on collective entrepreneurship—particularly the organizational and capital-acquisition aspects of innovative agribusiness firm, and contract agriculture—with attention to both theoretical and policy concerns. Agribusiness DOI 10.1002/agr