Assessing the Impact of Corporate Governance on Performance: A Case Study of Agricultural Co-operatives in Mato Grosso do

This study focuses on examining the relationship between corporate governance practices and the performance of agricultural co-operatives in Mato Grosso do Sul, Brazil. The primary objective is to address agency conflicts within co-operatives by investigating the extent to which corporate governance practices align the interests of co-operatives and their members. The research approach involved a document analysis using the bylaws and meeting minutes of 21 co-operatives, totaling 84 documents. The assessment of corporate governance levels was based on a comparison with the 40 recommendations proposed by the Brazilian Co-operatives Organization (OCB, 2016), categorized into five major areas. The study is grounded in agency theory, particularly focusing on monitoring and incentives as key aspects. The findings reveal that the majority of recommended governance practices are concentrated in the "management board" area, while the co-operatives' most adopted practices are related to the "members and general assembly" dimension, which primarily involves monitoring efforts. However, when considering the full set of OCB's recommendations, only two were fully implemented by the co-operatives. These results indicate a low level of corporate governance in agricultural co-operatives in Mato Grosso do Sul.


INTRODUÇÃO
Agricultural co-operatives are collectively-owned organizations formed by farmers to cope with market failures (Costa, 2010;Sexton, 1990;Staatz, 1987). It is a business model that seeks a fair and balanced world, with better opportunities for all (OCB, 2017). For the international cooperative alliance -ICA, cooperatives are people-centered and controlled by their members, being the profits reinvested in the company or returned to the members (ICA, 2018).
They are controlled by their members, to whom the benefits are equitably distributed based on their use (Barton, 1989). Their aim is to improve the social interest of the organization and the production of goods and/or services with economic efficiency (Antonialli & Souki, 2005).
These organizations, as argued by Bialoskorski Neto (1998), have ideological and doctrinal motives, and are governed in Brazil by Law 5764, 1971(Brasil, 1971.
The main aspect of this law is related to decision-making rights, and establishes the rule "one person, one vote". Based on this, each member hasthe right of only one vote, despite his/her capital share. Theco-operative'ssurplus is distributed proportionally to the volume of activities or transactions he/she made. In sum, the aim is to generate benefits and not profit.
Co-operatives differ from other organizations because the member is simultaneously owner, user, and consumer of the co-ops products, and the right of control is not linked to the right to the residuals (Costa, 2010). The generation of benefits is the primary goal and the main motivation that explains the participation of co-operative members (Barton, 1989). The author points out that such benefits can be understood as: i) the opportunity to buy from or sell to the co-operativeat fair and efficient prices; ii) cost reduction through economies of scale and co-ordination; and iii) the achievement of market power.
According to the Brazilian Co-operatives Organization ( Collective firms, as well as private ones, present divergences between the interests of the individualand the organization, which in turn generates conflicts in the relationship between the parties, representing a challenge for management. Jensen and Meckling (1976) argue that these conflicts result from the separation of ownership and control, characterizing agency conflicts. 1 Considering collective firms, Bialoskorski Neto (1998) asserts that it is impossible for co-operative managers to always act in accordance withmembers' interests, due to various reasons such as seekingto maximize their own interests, incompetence, or negligence. Cook, Chaddad and Iliopoulos (2004) indicate that these conflicts are also enhanced becauseco-operativemembers assume different roles, such as owner, member, manager, and user. Thus, conflict of interest is a cornerstone in co-operative management.
However, it is necessary that co-operatives are managed according to the interests of the co-operative members. The alignment of interests is the focus of corporate governance (Correia & Amaral, 2006) which is conceived in this research under the framework of agency theory. According to Silva, Souza and Leite (2011) corporate governance is one of the possible ways to align the interests and help to reduce agency conflicts.
Based on the assumption that conflicts of interest are one of the main problems faced by co-operatives, and that alignment instruments (in other words, the adoption of governance practices) contribute to the reduction of agency conflicts, we pose as a research question: What is the standard of corporate governance presented by agricultural co-operatives in the state of Mato Grosso do Sul (Brazil)? 2 Based on governance practices recommended by the "Guide of Corporate Governance" prepared by the Brazilian Co-operatives Organization (OCB, 2016), this study aims to analyze the adoption of corporate governance practices in agricultural co-operatives of Mato Grosso do Sul (MS).
The paper is structured in three parts, besides the introduction and the final considerations. The first part presentsthe theoretical assumptions of agency theory, the corporate governance concepts and practices, as well as the types of agency conflicts present in co-operative organizations. In the second part, the methodological procedures are discussed. The third part presents and analyzes the main results.

THEORETICAL APPROACH
The theoretical framework is developed in three steps. First, we discuss the main assumptions of agency theory and the problem of alignment. Next, the concepts of corporate governance are presented, focusing on their application to cooperative organizations. Finally, we analyze agency conflicts in co-operatives, and the role of corporate governance in minimizing agency problems.
Agency theory analyzes economic transactions under a contract perspective (Bialoskoski Neto, 1998), in which one party, called the principal, is responsible for hiring the other party, called the agent, to act in the interest of the former (Hendriksen & Van Breda, 1999).
However economic actors-the principal and the agent-have different levels of information. Usually the principal is less informed about the agent's behavior, enabling the agent to proceed according to their own interest. This situation can be characterized as information asymmetry; the agent provides incomplete information 6 to the principal, and this lack of information is the core aspect of agency relationship (Araújo & Sanches, 2005). Jensen and Meckling (1976) argue that individuals are rational and seek to maximize their individual well-being. It is assumed that the agent does not always act in order to meet the principal's interests due to the aim to maximize his own utility, ordue to incompetence, negligence, or bad faith. Hence, agents' decisions are costly because it is difficult and/or expensive to the principal to verify the agent's performance.
Jensen (2000) describes two situations with this problem. First, the agent keeps its position in the organization and follows goals that do not match the interests of the principal, resulting in reduced performance of the organization.
Second, the lack of incentives toward an effort of the agents' activities reduces the market value of the organization. Fama and Jensen (1983) depict some mechanisms to control the agent's behaviorwhere it does not align with the interests of the principal. Such mechanisms can be understood as a hierarchy of decisions, in which the lower levels of decisions require ratification and subsequently monitoring. Next, we have the replacement of the highest levels of management by boards of directors, ratifying and monitoring the most important decisions of the organization and the establishment of incentive structures of mutual motivationand monitoring. For these, control mechanisms (supervision/monitoring) and incentives are key to minimizingagency conflicts.
Corporate governance appears to be a way to reduce agency conflicts, mainly those related to information asymmetry (Carvalho, 2002). According to this author, corporate governance is a set of rules that aims to minimize agency costs, helping to align agents and principal interests.
The concept of corporate governance arose at a moment of a huge dispersion of capital ownership and a high degree of independence achieved by company managers (Dubeux, 2001). This author argues that the excessive freedom given to the managers began to be questioned, leading to discussions and debates about containing the activities of managers to a reasonable extent. It was necessary to adopt some practicesto minimize potential conflicts of interest that might arise (Nascimento, Bianchi, & Terra, 2006).Corporate governance is understood as a set of rules that leads the organization to the performance expected by the company, reflecting the rights and expected behavior of executives, shareholders, and investors (Haspeslagh, 2010).
According to Shleifer and Vishny (1997), corporate governance (CG) refers to a number of mechanisms by which resource suppliers guarantee the return on their investment. For Slomski et al, (2008) corporate governance is a system of decisionmaking and management practices that aims to determine and control the development and strategic direction of the corporations. Nascimento, Bianchi and Terra (2006) understand CG as the effort to align the goals of the individuals who make up the company's management with the interest of shareholders, through more effective monitoring. Colombo and Galli (2010) argue that corporate governance has gained importance because it investigates the alignment mechanisms of interests between different parties within an environment characterized by information asymmetry.
Theauthors also point out that the core idea is that the greater the degree of governance, the greater the alignment of interests between the parties. Thus, corporate governance aims not only to minimize agency conflicts (Renders & Gaeremynck, 2012), but also to influence the formation and the performance of companies in a developed economy (Vintem, 1998).
It is clear that the main concern of corporate governance is to create an effective set of mechanisms, both incentives and monitoring, to ensure that the managers' behavior is always aligned with the best interest of the company, minimizing conflicts that may exist in organizations.

AGENCY CONFLICTS IN CO-OPERATIVE ORGANIZATIONS
As already mentioned, agency conflicts arise when ownership and management are not embodied in the same individuals, especially when, in the presence of information asymmetry, both parties pursue self-interest (Jensen;Meckling,1976;Costa, 2010).
This phenomenon is at stake in co-operatives, favoring the emergence of various conflicts between the members and the co-operativeboard. In addition-since the member, the owner, and the user are the same person-challenging, complex, and recurrent agency problems arise in co-operative management (Cook et al, 2004).
It is therefore necessary to ensure that co-operativemanagers are in line with the interests of the co-operative members, which is the core aspect of corporate governance (Correia & Amaral, 2006). The adoption of governance practices is one possible way of aligning the principal (members) and the agents' (co-operative board) interests, contributing to the reduction of agency conflicts (Silva, Souza, and Leite (2011).
In co-op organizations, control is entrusted to a body of managers, creating potential agency conflicts. Silva, Souza, and Leite (2011) argue that agency conflicts may emerge in the relationship between the members and the board of directors, supervisory board, and executive managers. These dimensions are shown in Table   1.

Board of Directors
Members who are part of the board of directors are usually few, favoring the concentration of decision-making power in their hands; Because only co-op members can stand for election to the board, they may not always be prepared to exercise administrative functions; The mandatory renewal by election of at least one third of members of the board makes it possible that some of them are perpetuated in the direction of the cooperative; When the co-operative chooses to create administrative bodies with the hiring of independent managers, these managers are usually subordinate to the board of directors.

Supervisory Board
The responsibilities of the supervisory board are limited, as they are formed, at least in Brazil, only for co-operative members who often are not trained for this function; Although members of the supervisory board are not allowed to have family relationships with members of the board of directors, they can often have strong social ties, what facilitates collusion; The opinion of the supervisory board is generally associated with the results that are provided by the board, which are annual. Therefore,the supervisory board does not usually exercise effectively its supervisory function.

Executive managers
Executive managers tend to act in accordance with the board of directors, to which they are subordinated, and not necessarily in the interest of = co-op members. This situation is even more serious when the activities undertaken by managers are less transparent to the co-operative members. Source: Adapted from Silva, Souza and Leite (2011).
Besides the agency conflicts mentionedabove, there are other situations that also generate conflicts in co-operatives. Bialoskorski Neto (1994) points out that the member assumes different roles at the same time: owner, manager, user, and customer. Thus,aco-operative member could be both agent and principal of the same contractual relationship. This situation can lead to inefficiency, since this member can influence his or her own compensation as a manager, or decide in favor of the customer's interest and not that of the co-operative itself. Zylbersztajn (2002) agrees with this argument, stressing that this point is especially relevant since it enhances the level of complexity in a co-operative management.

CORPORATE GOVERNANCE AND PERFORMANCE
Theoretical and empirical discussions in the field of corporate governance seek to demonstrate that business organizations with good governance practices tend to present better performance (Silveira, 2004). However, the results do not yet present a unanimous conclusion about whether corporate governance is positively related to good performance (Ferreira, 2012).
On the other hand, some surveys present governance as an endogenous variable, which can be explained by corporate performance (Silveira, 2004). In this case, companies adopt such governance practices in function of a previous good performance, and thus performance can also influence and change the ownership structure of the company (Melega, 2011).
Specific to co-operative organizations, Pinto, Funchal and Costa (2008) evaluated whether governance variables impact the performance of Brazilian credit co-ops. According to them, there was a statistical relationship between governance standards and the performance presented by these organizations.
Based on the above-the relationship among corporate governance, agency conflicts,and economic performance-three assumptions are presented.

METHOD
This research focuses on the co-op bylaws and meeting minutes from agricultural co-operatives in the state of Mato Grosso do Sul, Brazil. This is an exploratory study with a qualitative approach, developed using the content analysis technique. Based on Bardin (2011, 83)  Each document was initially read by one researcher, and once a best practice was identified it was written down in spreadsheet in order to compile the data. A second researcher made the validation of the information collected, and later a comparative analysis of the generated data was made.  The recommendations of governance practices associated to each dimension are presented from Table 2 to Table 6 4 .

Table 2-Members and the General Assembly Dimension
Members R01 The co-operativemust adopt integration programs for new members.
General Assembly R02 There must be pre-meetings.

R03
The meeting callmust be available in advance to the Ordinary General Assembly.

R04
The notice must be published in different media, giving greater scope to members. R05 The meeting minutes must bereadily available to members.

R06
The meeting minutes must be electronically available in full to the co-operative members.

R07
The co-operative committees should have suggestions for topics to be included on the meeting agenda.

R08
The co-operativemust have some manual to stimulate and facilitate the participation of co-operative members in the assembly. R09 There must be mediation and arbitration mechanisms. Source: Elaborated by the authors

Management
Board/ R10 Executives must be elected or hired.

R11
The board of directors must not intervene in operational matters.
Directors R12 The advisors must seek specific certification with the recognition of OCB system.

R13
The chairman of the board of directors must ensure that counselor directors receive complete and timely information. R14 The chairman must not take up the post of chief executive.

R15
The performance of the board of directors and of counselor directors must be assessed. R16 The directors' compensation must be linked to mediumand long-term indicators.
R17 The board meetings must be in person.
R18 The meeting minutes must circulateamong the participants prior to their approval. R19 There must not be insider information to any co-operative member.
Executive Management R20 The board of directors must be separate from the executive board.

R21
The board must apply an intelligence system to track the co-operative's performance. R22 The board of directors must appoint an ethics committee.
R23 The fixed and variable executive compensation must be linked to results. Source: Elaborated by the authors

Members R24
The co-operative members must attend and participate in pre-meeting events and/or general meetings.

Supervisory Board
R25 Technical capacity must prevail in the supervisory board members.

R26
Members of the supervisory board must have no business with the co-operative, except asco-operative members.

R27
The advisers of the supervisory board must seek specific certification with the recognition of OCB system. Source: Elaborated by the authors

Table 5-Advisory Bodies and Audit Dimension
Technical committees R28 The co-operativemust adopt temporary committees whenever possible.

R29
The audit committee must regularly meet with the board of directors, the supervisory board, and the executive director.

R30
The ethics committee composition must meet the prerequisites of regulatory agencies. R31 The ethics committee shall be guided by secrecy.

R32
The human resources committee must assess matters relating to succession, compensation, and people development.

R33
The human resources committee must regularly meet with the CEO and the cooperative department in charge of people management. Internal Audit R34 The internal controls, standards, and procedures and the monitoring of internal audit must be improved.

Independent Audit
R35 The general assembly must be aware of the activitiesof the external audit firm.

R36
The independent auditors must not be hired to perform consulting work. Source: Elaborated by the authors

Ombudsman R37
The ombudsman must welcome the manifestations of the co-operative members. R38 The co-operative committee (OQS) must perform ombudsman services.

Relationship
with the members

R39
There must be communication between the governance structure and the members' board. R40 The co-operative members' satisfaction must be gauged. 14 In order to measure the level of corporate governance, an index presented by Carvalhal-da- Silva and Leal (2005) was applied. In this sense, the adoption of governance practices was measured through dichotomous variables, which refer to the presence or absence of governance attributes. Based on the comparison between the recommendations of governance practices identified in the guide of Co-Operative Governance elaborated by the Brazilian Co-operatives Organization (OCB, 2016), and the legal documents analyzed (minutes and by-laws), it was possible to calculate a Governance Index (IGOV) of between 0 and 40 points.
For each good practice recommendation, if the co-operative presented the governance attribute, it would receive a score of "1", and if not a score of "0". In this way, it was also possible to transform the IGOV score into coefficients measured by percentages. This procedure has been used by several authors like Carvalhal-da- Silva and Leal, 2006;Santos and Leal, 2007;Silveira et al 2007;and Kapler and Love, 2002. Thus, it was possible to identify the adoption of governance practices in real and percentage terms, as well as to order the co-operatives according to them.
Co-operatives that had adopted governance practices of less than 20% were included in the Lower Governance grouping. Those that obtained adoption levels between 20% and 25% were in the Medium Governance, and those above 25% were considered as the High Governance Grouping.
To test the assumption of a positive relationship between corporate governance and performance, specifically in co-operatives, we first reviewed the financial information of the sample agricultural co-operatives for a period of four years (2010 to 2013).
Due to the absence of some financial information during this time, four cooperatives were eliminated from the initial sample, and the final sample consisted of 17 agricultural co-operatives from the Mato Grosso do Sul.
As economic performance variables, we used the indicators calculated from the database shared by the Observatory of Co-operatives of the Ribeirão Preto campus of the University of Sao Paulo School of Economics, Business Administration, and Accountancy (FEARP / USP). These indicators are presented in Table 7. Shows the net revenue attained for each member. Source: Based on Santos (1986) The analysis of variance (ANOVA) for each indicator of economic performance verified whether, on average, its performance was the same for the differentiated levels of corporate governance studied. This analysis was performed for each of the years analyzed (2010 to 2013).
The null hypothesis tested is that the mean value of the indicator is the same for the different levels of corporate governance; that is, there is no influence of the level of governance found in the average economic performance measured. The alternative hypothesis indicates that there is some difference in the average economic performance for the different levels of economic performance. The level of significance (α) equal to 5% was used, so differences will be detected for P-value lower than 0.05.
Regarding the sample, it is composed of 21 singular co-ops (first-level). The oldest one has been operating for 81 years, and the youngest 5, with the average being 21. The number of members in the co-operatives varies from 26 to 947 associates. In general, the main activities are related to corn, soybean, sorghum, milk, and egg production. In order to compare the level of corporate governance among them, the co-ops were randomly numbered from 1 to 21.
The sample characterization is presented in Table 8. The co-operative characteristics include: (I) the county seat; (II) the coverage area of the operation; (III) the year of foundation; (IV) age of the organization; (V) the number of associates in 2015; and (VI) the main products.

RESULTS
This section is divided into threemain topics, as follows: i) Theoretical variables and governance recommendations; ii) Co-ops and corporate governance; and iii) Ranking agriculture co-ops according to the level of corporate governance.

THEORETICAL VARIABLES AND GOVERNANCE RECOMMENDATIONS
The guide of co-operative governance elaborated by the Brazilian Co-  The theoretical presupposition of this research is that the adoption of 5 Details about each dimension and its recommendations are presented in item "3" (Method) of this paper. corporate governance practices aligns the co-operative's and the members' interests, which in turn minimize agency conflicts. Thus, it is important to link theoretical variables to each governance practice recommended by OCB. The theoretical variables are: i) monitoring; and ii) incentives.
The majority of governance recommendations are related to monitoring efforts (62.5%), followed by recommendations designed to reduce information asymmetry (37.5%). Regarding the promotion of incentives, we identified 17% of recommendations. Thus, considering the OCB Guide, the focus is primarily on monitoring the agents' behavior.

CO-OPS AND CORPORATE GOVERNANCE
The level of adoption of corporate governance is analyzed by each dimension: i) Members and the General Assembly; ii) Management Board; iii) Supervisory Board; iv) Advisory Bodies and Audit; and v) Ombudsman and co-operative relationship with members.

i) Members and the General Assembly (MGA)
This dimension accomplishes nine recommendations, of which only one showed 100% adoption by co-operatives. This recommendation is related to the necessity of making the call for the assembly available in advance to the meetings.
The relevance of this recommendation, according to OCB (2016), is that the members can be properly prepared to attend the meeting and to deliberate about cooperative decisions.
On the other hand, three recommendations in this dimension are not adopted by any co-operatives, namely: i) the meeting minutes must be electronically available in full to co-operative members; ii) the co-operative must have some manual to stimulate and facilitate the participation of co-operative members in the assembly; and iii) there must be mediation and arbitration mechanisms.

ii) Management Board (MB)
In this dimension there are 14 recommendations of governance practices, and only one showed 100% adoption by co-operatives, which is that the board meetings must be in person. As stated in OCB guide, the other means the council could meet were only permissible in exceptional cases.
The other seven recommendations in this dimension were not adopted by any co-operatives, namely: i) the advisors must seek specific certification with the recognition of OCB system; ii) the performance of the board of directors and of counselor directors must be assessed; iii) the directors' compensation must be linked to medium-and long-term indicators; iv) no insider information may be provided to any co-operative member; v) the board must apply an intelligence system to track the co-operative's performance; vi) the board of directors must appoint an ethics committee; and vii) the executive director's compensation must be linked to results.
In general, it is observed that the performance and behavior of the co-operative management board is not assessed.

iii) Supervisory Board (SB)
In the Supervisory Board dimension, it was found that of the four recommendations, only two are widely adopted by the co-operatives. The first, with 86% adoption, is related to the recommendation of attendance and participation of co-operative members in pre-assembly meetings and general events. The recommendation is that co-operative members must monitor the accountability of management bodies, as they are the most interested party in the co-operative business and should act like instruments of effective monitoring and control.
The other recommendation, adopted by 38% of co-operatives, is that members of the supervisory board must have no business with the co-operative, except as a co-operative member, because according to OCB (2016), the effectiveness of the supervisory board performance depends on its independence On the other hand, the recommendations regarding the predominance of technical capacity of the members of the supervisory board, and the need for specific board certification recognized by the OCB system, have not been adopted by any cooperatives of this research.

iv) Advisory Bodies and Audit (ABA)
There are nine recommendations in this dimension. Two of them have been taken up by co-operatives in the study, each with a 57% of adoption rate. They are related to the recommendation to adopt temporary committees whenever possible, and to not hire as consultants independent auditors who have already carried out work for the co-operative. According to the OCB guide, the activities of temporary committees should be periodically evaluated to ensure its effective role, and not hiring auditors' firms which have already acted as consultantsavoids conflicts of interest.
Of all the recommendations of this dimension, five were not adopted by any of the co-operatives, namely: i) the audit committee must regularly meet with the board of directors, the supervisory board, and the executive director; ii) the ethics committee shall be guided by secrecy; iii) the human resources committee must regularly meet with the CEO and the co-operative department in charge of people management; iv) the human resources committee must regularly meet with the executive director and the co-operative department in charge of people management; and v) the internal controls, standards, and procedures and the monitoring of internal audits must be improved.
It becomes evident that practices related to audit and the creation of other advisory boards, besides the management and supervisory board, are not usually adopted by the co-ops.

v) Ombudsman and co-operative relationship with members (OCM)
The dimension related to the ombudsman and co-operative relationship with members presents a total of four recommendations, two of which have been adopted by twoco-operatives. These recommendations relate to: i) the ombudsman must welcome the manifestations of the co-operative members; and ii) the co-operative committee (OQS) must perform ombudsman services.
As suggested by OCB guide (2016), it is recommended to welcome the cooperative members' manifestation, such as compliments, suggestions, queries, and complaints in order to resolve any potential conflicts. Also, the co-operative must develop some sort of organizational structure to enhance its relationship with members, creating bridges between the governance structure and the membership, reinforcing the idea that the co-operative belongs to the members, sharing information with transparency, with a focus on the members' continuous education.

iv) Comparative Analysis of the adoption of Corporate Governance
According to Table 9, co-operative C16 was the one which adopted the greatest number of governance practices (13 recommendations) proposed by OCB Manual, which represents 32.5% of all recommended practices. Such practices are mostly related to monitoring agent's behavior. This singular (first-level) co-operative has a national scope and its main products are corn, soybean, cotton, and sorghum.
Only nineco-operatives in this study adopted more than nine recommendations, and only four show governance practices in all dimensions.
One third of the co-operatives adopted 22.5% of the recommendations of good practice. These recommendations are related to both monitoring and assuring that the information be spread among all stakeholders. These are all singular (firstlevel) co-operatives with local (2),state (4), and national (1) scope,and are on average 18.5 years old. They are mainly involved with the production of corn, soybean, sorghum, milk, wheat, cotton, pigs, and beans. On the other hand, in two co-operatives the adoption of governance practices was less than 14%, although they have around 500 members and are more than 20 years old.

GOVERNANCE
We ranked the co-operatives into groupswhich presented differentiated levels of corporate governance, according to the increasing degree of governance (Governance Index -IGOV) related to the adoption of governance practices. The coops were grouped into three levels of governance: "Low", "Medium" and "High". 6 The grouping had the objective of considering the difference in the average variation of the value of the indicators of each grouping by means of tests of variance (ANOVA). Table 10 shows the co-operatives divided into three differentiated levels of governance. Each co-operative is related to the percentage of adoption of good governance practices for the five dimensions of analysis used in this research. Those co-operatives that had adopted governance practices of less than 20% were included in the lower governance grouping, six in total. Those that obtained adoption levels between 20% and 25% were included in the group "Medium Governance", accounting for nine co-operatives. The six co-operatives with an 24 adoption level above 25% were classed as "High Governance". It is important to stress that "High Governance" is related to a comparison among the co-ops in this sample. In other words, the co-ops classified as "High Governance" in this research are those that had adopted more than 25% of the practices recommended by OCB.
As already mentioned, co-operative C16 had adopted the highest number of good practices recommended by OCB, at 32.5 %, which in fact is not really a high score itself.
As variables of economic performance, the indicators calculated from the database shared by the co-operative observatory FEARP / USP were used. These indicators, as suggested by Santos (1986), can be understood as: i) Profit Margin; ii) Turnover of Operating Assets; iii) Return on operating assets; and iv) Co-operative performance. Co-operatives that presented the necessary financial information for a period of four years (2010 to 2013) were sought, in order to lend greater robustness to the analysis. During this time interval, 4 co-operatives were eliminated from the initial sample, and the final sample consisted of 17 agricultural co-operatives from Mato Grosso do Sul.
Also, for the years 2010 and 2013 other co-operatives were eliminated because they did not provide the necessary information. For 2010, fourco-operatives were eliminated, and in 2013 another seven were eliminated. Table 11 shows the number of co-operatives for each level of governance after the elimination process, totaling 57 observations. its performance was the same for the different levels of corporate governance studied, using analysis of variance (ANOVA). This analysis was performed for each of the years studied (2010 to 2013) ( Table 12).
The null hypothesis tested is that the mean value of the indicator is the same for the different levels of corporate governance, that is, there is no influence of the level of governance found in the average economic performance measured. The alternative hypothesis indicates that there is some difference in the average economic performance for the different levels of economic performance. The significance level(α) equal to 5% was used, so differences will be detected for Pvalue lower than 0.05. Therefore, at least within the limits of this sample, one cannot confirm the influence of corporate governance on the economic performance of the agricultural co-operatives of Mato Grosso do Sul, thereby not confirming the third assumption raised in this research, which argues that the adoption of good governance practices contributes to a better economic performance of co-operatives.

FINAL COMMENTS
This study examined the standard of corporate governance practices adopted by agricultural co-operatives in Mato Grosso do Sul, Brazil, based on recommendations proposed by the guide elaborated by the Brazilian Co-operatives Organization (OCB, 2016). The theoretical presupposition is that the adoption of corporate governance aligns the co-operative's and the members' interests, thereby minimizing agency conflicts.
In order to answer the research problem, we addressed the following issues: i) the pattern of corporate governance recommended by OCB; ii) the analysis of the association between theoretical variables and governance practices; iii) the adoption of governance practices by co-operatives, by dimension; and iv) a comparative analysis among the co-operatives. For that, we investigated the bylaws and meeting minutes of 21 agricultural co-operatives, from 2012 to 2014, totaling 84 documents.
In general we conclude that the majority ofgovernance practices adopted by the co-ops are related to two main dimensions: i) Management Board (35%); and ii) Members and the General Assembly (23%). Also, the majority of governance practices are related to monitoring efforts (45%), followed by recommendations designed to reduce information asymmetry (38%), and to promote incentives (17%).
Considering the total of recommendations, only two had a 100% of adoption rate. The first of these is related to the necessity of making the call for the assembly available in advance to the meetings, and it is part of the dimension called Members and the General Assembly. The second recommendation adopted by all cooperatives is in the dimension Management Board, and it is concerned with the necessity that board meetings be in person. This recommendation benefits the effectiveness of the meetings.since the discussions are better developed with a deeper understanding of the relevant issues, also facilitating the analysis of the informative documents and encouraging their timely signing.
Considering all the recommendations of best governance practices, about 50% have not been adopted by any co-operative. Hence, we conclude that corporate governance in agricultural co-operatives in Mato Grosso do Sul has still a long way to go. It was also observed that monitoring is the main mechanism adopted by co-ops to align the members and the co-operative's interest, i.e. to minimize agency conflicts.
After applying the variance test (ANOVA) for each of the years studied, it was verified that only the profit margin and return on operating assetsindicators in the year of 2012 presented a level of significance lower than 0.05: that is, showing statistically significant differences in the average variation of the value of each indicator in the differentiated levels of governance.
On the other hand, in the years 2010, 2011, and 2013, the performance indicators presented values of significance higher than 0.05. Thus, it can be said that on average, in the three years analyzed, there was no significant difference in the values of the economic performance variables among the differentiated levels of governance.
In this sense, it is not possible to empirically confirm that the agricultural cooperatives of Mato Grosso do Sul with a better governance pattern perform better.
There is no improved economic performance by the co-operatives of the high governance group compared to the middle governance group, and middle governance in relation to low governance. Therefore, it cannot be inferred that good corporate governance practices affect the economic performance of agricultural cooperatives in Mato Grosso do Sul, at least within the limits of this sample.
For future studies, we suggest expanding the sample, incorporating a larger number of agricultural co-operatives from Mato Grosso do Sul and other Brazilian states. It would be interesting to conduct field research, with some face-to-face interviews, to better understand the level of corporate governance and the existence of informal practices. Finally, a future research agenda could incorporate quantitative methods in order to analyze the association of a better governance level with the cooperatives' performance. Moreover, it is important to balance the importance of each recommendation proposed by OCB, as in this research we did not assign different weights to each one.
We believe that this research has two main contributions. The first is to shed light on corporate governance practices in co-operative organizations in Brazil, which from our point of view is still in its first steps. Further, we identified and systematized at least 40 recommendations from the guide of Corporate Governance to Co-ops edited by the Brazilian Co-Operatives Organization (OCB). In sum, we present here a study that could be interesting either to the academia or to the private sector.
In spite of a possible criticism of fact that our analysis of adoption of governance practices was based on statutes and minutes of the meetings, which could potentially not reliably portray the practices adopted, we emphasize that if such official documents do not reflect the reality of the management of co-operatives, we have here a contribution of this research, which is to highlight the informality of the management process in co-operatives.
In view of the foregoing, as a suggestion for future work it is recommended to increase the number of co-operatives in the sample, and also to apply primary data, conducting field research to complement the documentary research, so that the data are confronted with the reality experienced by the managers. Anyway, this is an exploratory research and collecting primary data was not our first research design proposal. We conclude that "governance in co-operatives" can be treated in a more comprehensive way, thereby contributing to the expansion and improvement of studies that involve the corporate management of agricultural co-operatives.