Gender differences in agri-marketing farmer organizations in Uganda and Malawi: Implications for R4D delivery mechanisms

Although farmer organizations are acknowledged to link their members to markets, enhance their business skills and enable access to services, there is need for evidence that ascertains if all members benefit equitably. This paper examines benefit distribution and sources of gender differences in mixed-sex collective marketing farmer organizations in Uganda and Malawi. Through a cross-sectional survey, key informant interviews and gender segregated focus group discussions, data were pooled from 10 farmer organizations and 492 individual members. An independent samples t-test revealed that in both countries, men benefited more than women. In Uganda, significant differences existed in skills training, planning meetings, business networks, and impromptu general meetings. In Malawi, significant differences manifested in extension training and decisions regarding investment and marketing. In both countries, men were significantly more dissatisfied with the collective marketing initiatives than women, which resulted in side selling in Malawi. On the other hand, women were significantly more satisfied than men with the collective marketing and, in Malawi, they sold significantly higher quantities through their organizations than men. We conclude that collective marketing farmer organizations connect women members to markets but neither address gender inequalities nor stimulate women empowerment. We propose strategies that can improve gender responsiveness in mixed farmer organizations.


INTRODUCTION
Farmer organizations (FOs) have been advanced by several governments and other development stakeholders as a means for alleviating market access challenges that impede smallholder farmers from competing in agricultural markets (Mojo et al., 2017;Wanyama et al., 2008;World Bank, 2007). This is because through collective action, FOs can enable smallholder farmers to strengthen their bargaining power and participate in profitable value chains, as well as exploit economies of scale to transact business with big buyers of their outputs or suppliers of the inputs they need (Hilhorst and Wennink, 2010;Penrose-Buckley, 2007;Shepherd, 2007). Where extension service is limited or lacking, FOs such as cooperatives and farmer associations have been promoted as an efficient approach for providing support services e.g. extension advice, market information, credit and market outlets to farmers (Chagwiza et al., 2016;Otieno, 2019;Bijman and Wollni, 2008;Shepherd, 2007). In developed nations, such as United States, Canada, and Japan, agricultural cooperatives have become a powerful economic force shouldering marketing, supply, credit, insurance, as well as handling international businesses (Aref, 2011;Birchall, 2004). In Africa and other developing countries, cooperatives have been appreciated for creation of employment, poverty reduction, facilitating technology adoption (Bayan, 2018;Chagwiza et al., 2016), extending social protection, and representation of interests of the majority poor who have no voice (Develtere et al., 2008;Mrema, 2008). Other analysts find that cooperatives provide more inclusive trade relations and value chains and offer opportunities for gender mainstreaming and women empowerment (ICA, 2016;Thangata, 2016), helping poor women to lift themselves out of poverty and have their voices heard collectively (ILO, 2014). However, in all cases, authors recommend deliberate actions to increase and incentivize women participation in cooperative activities, leadership and decision making and enabling active engagement.
Despite the growing literature on the potential benefits that FOs give to their members, not enough studies have been conducted to understand whether the distribution of the benefits to members is fair enough to guard against empowering a few while using others as ladders of their success; thus, widening the poverty gap. The documented successes of FO as a mechanism for collective marketing have often not disaggregated results to indicate whether all members accessed the specific benefits that were generated by the FOs (Berdegué et al., 2008;Chirwa et al., 2005;Coulter, 2007;Kyazze, 2010;Ikwera and Twongyirwe, 2019;ILO, 2012;Shiferaw et al., 2009;Thangata, 2016;Wanyama, 2012;Mrema, 2008). Disaggregated information is particularly important for studies of mixed sex FOs, where women and men face unequal constraints that put women at a disadvantage when negotiating with men (Kaaria et al., 2016). Women are likely to join with the burden of triple roles, have low mobility and education relative to men, have limited access to and control over productive resources, and are generally expected to behave in a certain way by society (FAO, 2011). Thus, it is critical that disaggregated information on distribution of benefits among FO members differentiated by sex be made available to inform the design of strategies for addressing gender equity gaps.
This paper examines the distribution of benefits of participation in collective marketing among members of FOs and asks whether these FOs work equitably for men Ampaire et al. 917 and women members in the context of Uganda and Malawi. The paper contributes towards filling the gaps in literature on sex-disaggregated evidence in the context of FOs in different ways. First, we analyse the sexdisaggregated composition and governance of collective marketing FOs, which are of a mix of area cooperative enterprises (ACEs), also referred to as secondary cooperatives, and farmer associations at second-tier level. We focus on the sex of the members and occupants of leadership position as these may influence the nature and distribution of benefits. Then, we investigate the nature and extent of benefits that accrue to members and test whether there is equal distribution of those benefits between men and women. From the analysis, we draw implications for the potential of farmer organizations to deliver on gender equality and women empowerment as postulated by literature (ICA, 2016;ILO, 2015).

Historical perspectives of farmer organizations in Uganda and Malawi
In both Uganda and Malawi, FOs have been embraced and promoted as a means for rural development for over four decades. In Malawi, FOs started as farmer clubs in 1978 through the Ministry of Agricultural for purposes of facilitating farmer access to agricultural credit and extension advice in the country (Kishindo, 1988). These clubs have since evolved into a hierarchy of zonal/regional and national level apex organizations (Chirwa et al., 2005;Coulter, 2007;Diaz, 2004;Mapila et al., 2010a). In Uganda, FOs were started way back in 1913 as lobby pressure groups that mobilized to fight against the exploitation of European and Asian traders and get high returns from their marketed produce (Msemakweli, 2012;UNDP, 2016). At the time, the colonial government had enforced a policy that required smallholder farmers to produce cotton and coffee while the European and Asian business persons were required to process and market the products. Unlike in Malawi, cooperatives in Uganda served as a channel through which the government controlled marketing of coffee and cotton, which were important crops for foreign exchange earnings for the government of the time, while economic development and integrating the poor farmers into the economy became secondary. These cooperatives were linked to the market through the marketing board that handled most of the export marketing activities and provided credit in form of inputs to support production (Msemakweli, 2012).
Like many other countries across Africa, cooperatives in Uganda and Malawi also suffered from poor performance due to governments' interference that used them as platforms for rewarding political royalties with employment, thus running down their performance and reputation (Futures Agriculture, 2014). During this era, rules and regulations that came to shape the way cooperative members interacted were gender blind as women were only represented at the level of household headship while most women in agriculture living in male headed households were represented by their husbands.
After market liberalization that started in the 1990s, FOs underwent reforms that brought about changes in how FOs are organized and managed internally. They have since taken up multiple functions such as advocating for policy change (FUM, 2010), and serve as platforms for extension service delivery, mobilization of financial resources, agro-processing and collective marketing, as an avenue for linking members to markets (Biteete, 2019;Thangata, 2016).
The management of FOs in the post liberalization period has been highly decentralized in nature, which allows them to form their rules and procedures that provide a framework of incentives that shape their economic, political, social organization and behaviour (Dorward et al., 2009). According to the new institutional economics theory, this enhances the supervision of contracts and generates a stream of benefits to members that includes reduced transaction costs, risk, as well as overcoming problems of weak enforcement of laws and low community level social capital (Dorward et al., 2009). However, as argued in the Futures Agriculture (2014), organizational rules and procedures that shape incentives to members who cooperate are functions of internal and external factors mediated by gender aspects. Internally, power relations and group characteristics influence bargaining and decision-making processes, with more powerful members likely to influence the outcome of the interactions in their favour as the weak lose out (Kaaria et al., 2016).

Gender inclusion in the organization of FOs
The diversification of functions combined with decentralised managerial structures enabled participation of other social groups that were originally excluded from cooperatives during the pre-liberalisation era. In particular, social inclusion in farmer organization in the two countries shifted in favour of women. For example, FOs formed under the National Agricultural Advisory Services (NAADS) in Uganda, the dominant advocate for the group approach, were either constituted by more female members than males (ILO and ICA, 2014) or there were no differences (Benin et al., 2007;Davis et al., 2007). Different strategies have been promoted to enhance women participation in FOs. In the context where gender inequalities are deeply entrenched and socio-cultural norms prohibitive, the strategy has been to promote women only FOs (Carr and Hartl, 2008;Hilhorst and Wennink, 2010;Quisumbing and Pandolfelli, 2009). In case of collective marketing where the primary objective is linking farmers to new markets, women-only FOs are viewed as less successful in terms of market search, access to new information and negotiating with buyers (Barham and Chitemi, 2008;Gotschi et al., 2009;Katungi et al., 2008), which necessitates presence of men members in the organization to fill in the gaps. For these reasons, sex mixed FOs have tended to dominate where the interest is to link smallholder farmers to globalized markets through formalized FOs. Kaaria et al. (2016) highlight additional strategies that have been used to increase women participation in FOs. They include basing membership on individuals rather than the household; aligning entrance requirements to what women can control; reducing membership fees for women and getting more women on the executive boards. Authors caution that these strategies serve to bring women into a mixed organization but do not automatically guarantee their active continuous engagement.
In both Uganda and Malawi, women continue to be under-represented at political, policy and decision-making levels in mixed FOs. This is partly due to cultures that assign leadership and decision-making roles to men and organizational rules that attach high preferences on literacy skills as eligibility criteria for leadership (Mapila et al., 2010a), but discriminate against less educated members, majority of whom are women (IFPRI, 2005b;Mapila and Makina, 2012;Sinyolo and Mudhara, 2018).
Despite the pre-entry differences in capabilities, all FOs in this study adopted the voluntary and open membership cooperative principle: all persons who can use the FO services and are willing to accept the responsibilities of membership are free to join. Thus, men and women are subjected to the same criteria that qualify them as members -payment of membership and annual subscriptions fees and willingness to market a proportion of their produce through their FOs. The amounts of membership and subscription fees are democratically agreed upon and payable by all members. The members are also bound by a common set of rules that structure their behaviour, prescribe the rules of business conduct and a leadership code (Ampaire et al., 2013b). In mixed sex FOs, the leadership code stipulates the proportion of men and women that constitute leadership committees, among other things. The quota system, which is adopted from government macro policies as a blue print, is adhered to for women representation on executive leadership committees but not necessarily on subcommittees.
There is evidence that FOs can enable women to break through the community safety concerns or socio-cultural norms and access markets, increase their assets and gain control over decision-making processes that affect their lives (Hilhorst and Wennink, 2010;Penrose-Buckley, 2007). However, it has also been shown that the success of FOs in providing benefits to women members, particularly in mixed sex FOs, does not come freely but requires design and implementation of measures that address the constraining gender norms (Kaaria et al., 2016;Lecoutere, 2017;Quisumbing and Pandolfelli, 2009;Rubin et al., 2009). For example, there is evidence that successful agriculture production depends on whether one has access to enabling assets such as natural, physical, human, financial, and political capital (Meinzen-Dick et al., 2013). As such, the capacity of women to produce and improve livelihoods is constrained by lack of or limited access to and control over assets such as land, inputs, credit, extension services (Ampaire et al., 2013a;Lodhia, 2009;Rubin et al., 2009). Because women join FOs with such constraints, there is always need for FOs to put in place mechanisms that can help women overcome these constraints to not only be members of FOs, but beneficiaries of the benefits that come with it.

Data sources and collection methods
The data used in the study were obtained from both secondary and primary sources to enable triangulation of information during analysis. In Uganda, secondary data was obtained from documents provided by Uganda Cooperative Alliance (UCA), the apex body coordinating cooperatives in the country, who also recommended a list of ACEs to work with, from which a final sample was purposively selected based on the target commodity and ease of access. Additional documents were downloaded from the ministry of trade, industry and cooperatives website. At ACE level, transaction records were obtained from the executive, who also provided an updated list of member grower cooperative societies (GCSs). Membership records were obtained from the leaders of the randomly sampled GCSs. In Malawi, the beans programme at Kitedze research institute, in collaboration with NGO partners, sampled the five target farmer organizations. Leaders of the sampled associations provided business records and lists of their member clubs. Club leaders provided lists of their respective members.
Data from primary sources were gathered via a mixed methods approach combining gender segregated focus group discussions (FGDs), key informant interviews and a survey of individual FO members selected from primary cooperatives or clubs. Sampled FOs include five ACEs in Uganda, which were constituted by GCS and five second-tier farmer associations in Malawi that were constituted by clubs. The ACEs and associations were purposively selected to include those that prioritized bean production 1 as one of the priority commodities and collectively marketed the produce through the FOs at least once. Confirmation was made with the leadership of the organizations to ascertain that the FOs were still active and involved in collective marketing before the study commenced. A total of 22 GCSs and 26 clubs were sampled from Uganda and Malawi respectively. For each of the selected GCS/club, a full list of individual members was obtained from the leaders and stratified proportionate sampling method applied based on the composition of men and women in the total membership.
In Uganda, two FOs were selected from the eastern region, two from the central and one from western Uganda. Most FOs sold bean grain to local markets and only one FO supplied bean seed to a seed processing company through contract farming. Table 1 indicates that the socio-economic context of the study areas in Uganda can be distinguished in terms of socio-cultural dynamicsaccess to services and markets, and gender. In general, male dominance existed in all sampled communities although it is moderated by social networking behaviours. A total of 23 FGD constituting 184 farmers; five for top executive and 18 for members aggregated from 22 primary cooperatives in five ACEs were organized and conducted. The membership in secondary cooperatives in Uganda was spatially spread, demanding that we conduct more FGDs to take care of variability from 23 FGDs.
In Malawi, the study was conducted in Ntchisi, Dowa and Nkhotakota districts of the central region that is majorly matrilineal (Berge et al., 2014). This matrilineal lineage system offers the primary land rights to women which are passed on from mother to daughter (Kishindo, 2010a). The social economic-context of Malawi in which the FOs is embedded varies in terms of socio-cultural dynamics, type of collective businesses and support from service providers ( Table 2). Three out of five FOs sampled were registered under the grain legume association (GALA) umbrella, one association was supported by World Vision and one cooperative was registered by the ministry of agriculture and supported by Total Land Care. In all cases, beans were an additional collective enterprise to soybean, groundnuts, and maize although a random list of crops grown by the sampled households put beans in second position to maize. A total of 124 farmers were met in 15 FGDs, five for top executives and 10 for members selected from 26 clubs.
FGDs were administered by the principal researcher both in Uganda and Malawi, with support from translators and a voice recorder. Information gathered from FGDs was intended to understand FO history, activities, mode of establishment, membership, how farmers are linked to markets, markets supplied, sales, leadership capacity, governance structures, physical assets owned, access to other market support institutions such as credit, market information and extension services. They also made suggestions on how women participation and access to benefits could be improved. FO members were also asked if they received any benefits from their organizations and responses were captured as yes/no. Those whose response was yes were required to list benefits received ( Figure 1). The interviewer went ahead to probe for the specific benefits (Table 3) using a pre-scribed list that was generated from responses by the FO leaders. For decision makingand meetings, respondents listed decisions and meetings they were involved in the previous 12 months (Table 4).
Information from FGD was complemented by primary data collected from key informants and a formal survey of FO members. A total of 12 key informants were purposively selected to include ministry officials, extension staff and local leaders in both Uganda and Malawi. Key informants highlighted the history and relevancy of FOs at national and local levels, described the socio-economic and cultural context, and ascertained activities of the sampled FOs. For the survey, we adopted the recommendation by Sekaran (2003) for an appropriate sample size larger than 30 and less than 500 and sampled a minimum of 30 members per GCS/club depending on membership size. A total of 492 individual FO members (303 men and 189 women) from both Uganda (245; 167 men and 78 women) and Malawi (247; 136 men and 111 women) were interviewed in the cross-sectional survey. A team of trained enumerators with good knowledge of English and the local language conducted individual interviews using a pretested questionnaire administered through face-to-face interviews. The individual interviews elicited information on demographic characteristics of members, such as gender, age, position in household, marital status and education level. Information was also gathered on size of land holding, distance from residence to the bulking centre, commodities produced and those marketed, non-farm sources of income, housing structure, physical assets owned, participation in decision-making processes and meetings, access to and type of benefits generated by FOs and perceptions of satisfaction with FO performance. Only variables relevant to collective marketing were analysed for this paper.

Characteristics of sampled FOs
Tables 1 and 2 indicate that most sampled FOs in both countries were started purposely for collective marketing reasons. Additional roles such as value addition, bulk purchase of inputs, pooling savings for credit, provision of market information, transportation of produce and improving member access to services were added to strengthen collective marketing.
In Uganda, two out of five FOs had beans as  the priority enterprise while three ranked beans as a second or third priority crop. In Malawi, three out of five produced beans as the priority enterprise while two ranked beans in second or third position. However, beans remained a major enterprise due to availability of good local markets and its importance for both food security and income. At the time of data collection in Malawi, men were increasingly taking on the bean enterprise as a substitute for the 'staggering' tobacco enterprise. In both countries, we neither found statistically significant differences associated with other enterprises nor between bean seed and bean grain.
FOs in Uganda provided more services to members than those in Malawi (Tables 1 and 2). Probably this is because the FOs in Uganda have existed longer than those in Malawi. Observations and key informant interviews indicated that Malawi FOs had not received sufficient support regarding institutional building although all of them had received external support at the start.

Analysis of FO composition
The FO composition is analysed with respect to selected demographic and socioeconomic factors that have been reported to be important in decisions regarding participation in agricultural marketing by smallholder farmers since these FO were primarily initiated to enhance market access. Marketing through FOs offsets issues of high transaction costs associated with long distances from the market and constraints emanating from lack of marketing experience and market information. This means that individual farmers experiencing these constraints are likely to be attracted to FO but there might be differences between men and women. Table 3 presents key statistics of FO members in both Malawi and Uganda and gives a statistical comparison of selected characteristics between men and women members of the studied FOs in each country. Overall, results show that in both countries, the FO is composed of members' representative of a typical farmer in each context. A typical member is in his/her middle age (that is above 45 years) with few years of formal education. The average distance from home to the bulking centres is 2-2.4 km, which demonstrates that, members in FOs are recruited from close neighbourhoods.
Results reveal greater similarities between men and women members in Uganda FOs. In this context, both men and women were of similar age, come from households that have comparable size of land holding and location with respect to the bulking centres. This means that both men and women had to travel same distance to centralized aggregation centres from where the FO leaders transported the produce. Comparable land sizes do not mean that women own land but suggest right of access to use (Kabahinda, 2017), which is a very important factor that determines women's commitment to collective action through cooperatives (Selhausen, 2016). However, men were of higher education than women and had longer experience in FOs than women -thus at an advantage when dealing with challenges that come with collective marketing (Barham and Chitemi, 2008).
The story was slightly different in Malawi, with results showing more differences between men and women members of FOs that favour men over women when it comes to negotiations (Mapila et al., 2010a). In addition to possessing higher education and the tenure of membership in FOs, men in Malawi were on average older than women. Similarly, men were from households with relatively larger land holdings compared with women and reported significantly larger household sizes.

Governance of farmer organizations in Uganda and Malawi
Here, we examine the structures of leadership and decision-making processes.

Leadership in FOs
In both countries, FOs are generally governed through Ampaire et al. 923 democratically elected leadership committees. At the second-tier level, the executive committee is constituted by representatives from the member primary groups or clubs. Each organization is managed through the constitution that mandates 30% of the positions on the leadership committee to be held by women. This representation was adopted from government legal provisions as a blue print. Most FOs have additional subcommittees e.g. for marketing, procurement, monitoring), which offer additional leadership opportunities. These committees are small and tend to be technical thus do not often observe the 30% representation. The FOs are also managed through a set of written rules and regulations, enshrined in their constitutions. Other than the 30 percent women representation, the organization constitutions did not have in place any strategy that provide for gender inclusion or active participation of women. FO members in Uganda were asked why women participation in leadership is limited and 22% of responses pointed to women's multiple roles in the household, 18% argued that the FO constitutions are already predetermined that women participation is limited to 30 and 18% pointed to women being shy and lacking confidence, while 15% blamed it on women's low education levels. In Malawi, some of the FO leaders, especially men, neither appreciated the importance of involving women in leadership nor ensuring that they actively engage. In one FO, which had no woman representative on the executive committee (except on paper), the chairperson wondered why any attention should be paid to women. 'Rural women in Malawi are shy and lack confidence to take up leadership roles', he affirmed his stand. In one ACE in Uganda, a female treasurer explained that male leaders would in some cases not invite her for a meeting if they felt that she would not support their decision.

Decision making
Decisions in FOs are made as a consensus through meetings that are constituted differently and are held at different levels. At both the second tier and GCS/club levels, respective executive committees make day-to-day management decisions, based on the mandate laid out in the constitution. In other instances, e.g. abrupt requests from apex organizations that need backing by members and election of leaders, all members meet to decide. More technical decisions such as planning, problem solving, and marketing are made by the executive or assigned sub-committees. Table 4 compares men and women member participation in FO meetings and decision making, which are key indicators of democratic governance.
Results show that the participation of men and women in meetings is comparable in Uganda. In Malawi, significant differences between women and men Percent response has been calculated as a fraction of the total numbers of men and women that attended specific meetings or participated in making specific decisions. Asterisks *, **, *** denote significant at 10, 5, and 1% confidence levels.
participation exist regarding planning meetings, technical training and special general meetings that are often not previously planned for but rather are conducted impromptu to address specific issues that come up. Women's limited participation in planning meetings points to low representation of women in the leadership since these meetings are mainly attended by leaders. Their low education levels constrain attendance of trainings while inability to attend impromptu meetings signifies women's limited mobility. These findings agree with previous research by Kaaria et al. (2016). The observations show that FOs are not building in mechanisms to enable equitable participation of both men and women, which is a pre-requisite to equitable access to benefits. Between countries, both men and women in Uganda were involved in making most decisions with a few exceptions. The bivariate analysis shows that significantly fewer Ugandan women participated in making investment decisions while in Malawi, there were limited investment decisions made either by men or women. Significantly fewer women in Malawi were involved in making marketing decisions such as selecting the marketable enterprise and deciding who the buyer of their produce would be. Inability to involve women in marketing decisions reinforces a disempowered status quo where women's voices are not heard, their priorities are not considered on the FO agenda, and their confidence to undertake marketing roles remains low.

Sex-disaggregated access to benefits generated by FOs
In organizing for collective marketing, FOs strategize to benefit members in two main ways. They provide markets for members' produce in response to their primary objective. This is made possible through searching for market information e.g. prices that help in determining suitable buyers. Secondly, FOs provide other services or attract development stakeholders that provide the services that FO members would not be able to access as individuals. Such services include extension advice, value addition, grading and transportation of products to the market, and mobilizing savings that translate into low interest financial loans for members (Tables 1 and 2). The services offered by FOs and partner service providers are supposedly accessible to all members, who perceive these services as benefits that accrue to them because they are members. In the survey, individual members of FOs were asked to specify the benefits that they had received from the FOs they subscribe to and Figure 1 presents the individual responses provided by men and women. In both countries, most men and women FO members received agricultural extension services, agricultural inputs, and market information from their FOs. On average, a lower %age of women, compared to men respondents, accessed these benefits except in the case of credit and agriculture extension in Uganda. More Ugandan men and women FO members, compared to Malawians, developed personalized business networks from the connections initiated by the FOs. In general, fewer farmers in both Uganda and Malawi accessed credit and participated in management training. We tested these observations to see if any significant differences exist between men and women and results are reported in Table 5.
In Uganda, men had significant access to the management training compared to women. We analysed this further to find out the level of participation in the management training among FO leaders and members and found that out of 39 leaders that obtained training in Uganda, 32 were men and only seven were women and the difference was statistically significant (p=0.0001). This represents 15% of the total leaders that benefited from training which is far below the 30 % mandatory representation, which is evidence that women benefited less. In Malawi, five men and zero women leaders had attended the training, and the difference was statistically significant (p=0.084). Focused discussions with leaders revealed that the design and delivery of the management training was marred by inequalities. First, few women sit on the executive thus only those few can be trained. Second, the training is residential in centralized locations meaning that most women would be constrained from attending by either spouses or cultural norms (Kaaria et al., 2016). Third, the topics (group leadership, financial management, book keeping) are technical requiring literacy skills that women may not have, and the mode of instruction requires participants to be able to read and write. We conclude that in both countries, training of leaders does not only favour men but also reinforces existing gender inequalities.
Similarly, the training of ordinary members benefits men more than women. In Uganda, out of 61 ordinary members that received extension training, 42 were men and 19 were women. In Malawi, training on good agronomic practices, quality management and record keeping favoured men. Significant differences also existed in developing business networks in Uganda and access to extension training in Malawi. Women's limitation to forge business networks point to restricted mobility (Pandolfelli et al., 2008) and socio-cultural norms that restrict women's interaction spaces (Kaaria et al., 2016). The limited extension training by Malawian women could be the effect of the extension delivery which follows a unitary household model where the household head is invited to attend the training or receive inputs, among other things.
In both countries, members used FOs in addition to other market outlets. On average, 48 and 49% of marketed bean was sold through FOs in Malawi and Uganda respectively, which reflects the extent of member satisfaction with FOs. Women were significantly satisfied with FO performance in both Uganda and Malawi and they marketed a significantly bigger proportion of their marketable products through the FOs in Malawi. On the other hand, men were significantly dissatisfied with FO performance in both Uganda and Malawi. This is evidenced by the fact that men sold significantly lower proportions of their marketed produce though FOs in Malawi. About 23% of Ugandan respondents mentioned that the dissatisfaction stemmed from delayed payments that discouraged some members from collective marketing. In some ACEs, this dissatisfaction had resulted in factions in which some primary cooperatives started own bulking initiatives disputing the principle of selling collectively at secondary cooperative level. FO members were asked to list the constraints they faced in bulking. In Malawi, 36% of respondents said they lacked alternative sources of income to deal with emergencies, 19% said they harvested small quantities and 10 % were constrained by FO quality requirements.

FO composition and governance
Results from the study reveal significant difference between men and women members of FOs in Uganda and Malawi. In both countries, the results suggest that men had opportunities that equip them as better negotiators than women. Men had stayed in the FO longer, had better education and benefited from training, which makes them better equipped with skills and knowledge that women lack. This provides an advantage to women: as men pull new information into the group consistent with findings in Katungi et al. (2008) that men participated in information exchange as suppliers of information while women were likely to participate as recipients of information. This implies that FOs are beneficial in helping women to overcome marketing challenges that often constrain them individually e.g. small volumes of production, limited mobility and business connections, and low negotiation power (Ampaire et al., 2013b;Gotchi et al., 2009;Katungi et al., 2008). As a matter of fact, women demonstrated better satisfaction with the performance of FO collective marketing initiatives than their male counterparts because according to them, FO offers better prices than they would have obtained if they handled marketing outside the FO.
However, possession of better education and experience by men means that they can easily dominate leadership positions and decision-making processes in the group to the disadvantage of women if no deliberate interventions are designed and implemented. In fact, most of FO were structurally gender-blind and they select leaders based on criteria that does do not sufficiently consider gender gaps. As a consequent, women were disproportionately underrepresented at higher positions of leadership. Coupled with low technical capacities, women participated less in extension advisory, planning and decision-making meetings. These observations suggest that women may not have sufficient space to voice all their concerns and articulate their priorities on the FO agenda. Thus, they are unlikely to influence changes that address the limitations they come with into the FOs.

The quota system for women representation in leadership
Results have indicated that FOs in both countries Have adopted the quota system, but as a blue print from the government. Whereas the quota system serves as an entry opportunity to uplift the status of women, it is not sufficient for addressing all gender inequalities in collective marketing initiatives. Hence, there is need for other guidelines to ensure active participation of the few women representatives. However, none of the sampled FO had such guidelines for gender considerations. In the short term, the sub-committees offer additional opportunities for expanding women representation in the leadership if qualifications can be adjusted to admit more women. In the long term, FOs need to develop structures that enable gender equity. For example, Manchón and Macleod (2010) find that the National Federation of Cooperatives in Nicaragua increased women representation on the board from one to three and instituted a women's commission responsible for decision making and co-ownership of land to be able to overcome gender disparities.

Training in management, extension and business networks
Results revealed unequal distribution of access to training in management between men and women leaders of FOs. For example, in Uganda, more men than women trained in management and developed business networks than women while in Malawi, more men than women accessed extension training. The limited training in management and extension pushes the women plight further to the edge since they already have low education levels. Yet women's access to education, training and information has been found to be one of the key factors that influence women participation in producer organizations (Kaaria et al., 2016;Selhausen, 2016), and constitute important elements of empowerment (Meinzen-Dick et al., 2019). Thus, the inequalities do not only constrain women's participation but also impend their empowerment. This calls for changes in the extension delivery mechanisms to integrate gender responsive approaches such as the gender action and learning system approach.

Access to income from beans and credit
From FGDs we found that utilization of income generated from the sale of produce varied according to cultural values upheld by different communities. For example, in western Uganda, utilization of income from bean sales was jointly agreed in the household. However, in male dominant cultures of Eastern Uganda, income from beans was handed over by men to their wives for safe custody, but the latter had no authority to spend without the husband's permission -meaning that the husbands control the use of income. In both Uganda and Malawi, income generated by women from beans was mainly spent on household needs. In contrast, income generated by men was used for investments. With men controlling incomes, it implies that women may be shouldering the burden of feeding and attending to the details of household needs without income from beans. Based on findings of other studies (Blackden et al., 2006;IFPRI, 2005b), women's loss of control over income from beans also might lead to food and nutrition insecurity at household level since beans are such an important crop for nutrition security. Yet as Figure 1 indicates, few women have access to credit from FOs that they would use to bridge the gap. Further analysis indicated that only 24% of men and 20% of women in Malawi have just enough saving to meet their regular consumption needs. As some studies have indicated, lack of control over income, lack of access to credit and associated decisionmaking, signal to the fact that FOs in the study areas have not sufficiently contributed to empowering women to overcome the inequalities that constrain them from accessing markets (Sraboni et al., 2013).

CONCLUSION AND RECOMMENDATIONS
The paper sought to examine the distribution of benefits of members of FOs to find out whether men and women were equitably benefiting from being members of these organizations. Findings of the study have indicated that there are gender differences in the distribution of benefits; with men more likely than women to scoop higher benefits. The inequalities stem from the fact that men and women join the FO with differences in key factors such education, access to and control of production resources, which influence access to benefits. Despite these differences FOs do not build in mechanisms to mitigate barriers faced by some members to ensure equitable participation and access to opportunities by all members. The rules governing recruitment of members, selection of leaders and access to benefits treat members as a homogenous group. This results in women being underrepresented in the leadership, planning and decision-making processes. This means that increasing the number of women members in FOs neither guarantees them access to accruing benefits nor empowers them. Deliberate strategies need to be put in place to address contextual causes of gender inequalities. Based on the findings of the study, we recommend strategies that can redress the situation.

Implications of research for development delivery mechanisms
The study findings have demonstrated that FOs enable women to sell at better prices and access information via social network with men. However, there are still gaps to be closed. It is important that FOs increase women involvement in planning and decision-making processes and increase women access to skills training and extension service. The increased numbers of women in activities should be supplemented with guidelines that specify affirmative actions. FO constitutions can include statements that stipulate minimum numbers of women that attend extension and management training; and set a quorum that must be present before a meeting can take place or a decision can be made. Such guidelines can be developed by FO members with support from extension service providers and the apex organizations the FOs are affiliated to. However, service providers supporting these FOs should first create awareness and educate both men and women on the benefits of joint decision-making at both household and FO levels. At coordination level, umbrella organizations to which FOs are affiliated need to revisit the policies that guide formation and operation of FOs to make gender analysis and inclusion a priority in the FO business.

Increasing women's participation in FO leadership
FOs need to interpret the quota system as a minimum requirement but not a sufficient condition for women representation on leadership committees. Women representation should be increased beyond 30% on executive committees and should be extended to subcommittees, and their roles should be diversified beyond the treasury. There is also need for targeted capacity building to enhance women skills in management, decision making, networking and negotiation. In addition, FOs should modify requirements for women leaders from emphasizing literacy skills to include other capacities such as ability to speak in public. Writing skills should be mandatory only for secretaries and those responsible for record keeping but not others.

The need to improve women's access to credit
The findings made clear the important need for women to access credit. In some communities in Malawi, there were village saving schemes and FO members were encouraged to join. In Uganda, there are SACCOs affiliated to each ACE plus other merry-go-rounds.
However, women save small amounts, yet the amount of loans requested is often based on amounts available in saving. The small amounts accessible by women cannot be used for meaningful investment in agriculture due to competing demands such as children school fees, medical services and other domestic needs. FOs need to engage with private sector financial institutions to demand financial packages that are suitable for the different categories of members. Some ACEs in Uganda have been successful in accessing such loans from microfinance institutions as associations and these loans are distributed to members based on ability and size of business. Deductions for loan recovery are made by FO leaders every after sale and farmers receive balances. If gender responsive financial packages are negotiated and integrated in such initiatives, they can benefit women.

Implementing gender responsive extension delivery mechanisms
The current extension delivery mechanisms are largely gender blind. Gender is often understood as the numbers of men and women participating in an activity. There is need to move beyond representation to critically analyse and address issues that constrain women, men and youths from actively engaging in FOs. In Malawi, there is need to change extension service delivery from the unitary household model to the collective bargaining model to take care of competing interests and preferences in a household. Such shifts demand that extension service providers have gender analysis skills and technical capacity to educate the farmers and FO leaders about the need for gender responsiveness in day-to-day activities. There is apparent need for service providers and FOs to strengthen gender sensitization and community education about the importance of equitable involvement of men and women. An example that education of farmers is important is the Tikorane Project in Chiseu community, Kasungu district. This community went through the Enabling Rural Innovations (ERI) training in 2003/2004 (See a detailed description of the ERI approach in Kaaria et al. (2007). This community had deeply entrenched cultural barriers that kept women in the background e.g. they were never allowed to speak in public, were never believed to be leaders, and would not question husbands on anything including marketing produce and how the income was spent. After the ERI exposure, farmers responded by forming a bean seed producing association that is led by 50 % men and 50 % women representation and is guided by other gender responsive rules. Women are treated as equal decision makers both at the FO and household levels and are given extra time to voice their concerns in association meetings. At the time of data collection for this study, men still held the responsibility of marketing the bean seed, but women were very engaged and committed throughout the process. The women FGD revealed that the decision to sell at household level is joint, they receive the payment in the absence of their husbands, and money accruing from sales is spent as agreed by spouses. This has kept the women very motivated to invest energies in the marketable enterprise; they were involved in securing planting seed, collective spraying of member fields, calculating production costs, setting selling prices, and preparing produce for the market when buyers give notification. This example shows that institutionalizing gender norms in mixed FOs works for both men and women even amidst deeply rooted gender biased socio-cultural norms.

Limitations of the study and future research
The study focuses on analyzing benefits that accrue to members but does not consider other contextual factors that influence how farmers organize. Further research should analyze the contexts to come up with new innovations that can enrich gender inclusion in FOs. In addition, the study was based on bivariate analysis and was unable to establish casual relationships between the contextual factors with access to benefits, which is important to inform selection and design of interventions.
Although the lessons from this study are important in terms of informing research for development initiatives, the sampled FOs are not necessarily a representative sample of all FOs in the study countries. Since a variety of development agencies have undertaken supporting FOs, a more representative picture could be obtained by looking at a wide range of FOs across a diversity of enterprises. Innovations about what works for gender inclusion and women empowerment within the confines of FOs can better be reached by analyzing the different strategies that have been employed by different FOs and what worked well in which contexts. Future studies could focus on that.