Scaling-up: Gender integration and women’s empowerment in Southern Ethiopia

In the last couple of decades, the scaling up of successful pilot projects has been considered a crucial development strategy. The majority of scaling-up programs in developing countries stipulate the integration of gender as a central objective. In the article we argue that while the integration of gender in the scaling up of pilot projects has the potential to empower women, care should be taken not to overly focus on a segment of the women category (particular female heads of households) and overstate temporary gender gains that do not transform exploitative gender norms and practices. The article draws on evaluation research undertaken by researchers of the Scaling-up of Pulse Innovation for Food and Nutrition Security (SPIFoNS) Project, implemented in Southern Ethiopia. Data of this write-up were gathered using semistructured questions, focus group discussions and observation. The article argues that if projects such as SPIFoNS are to adequately challenge patriarchy and contribute to bringing about gender equity, they need to recognize the heterogeneity of women and design multidimensional programs that can help married women to gain full access to resources and participate in important household decision-making processes. Subjects: Development Studies; Gender & Development; Development Policy


PUBLIC INTEREST STATEMENT
This article draws on the experience of SPIFoNS project and highlights the challenges of integrating gender in a development programs to empower women. The article argues that in order to transform exploitative cultural values and practice, projects such as SPIFoNS that seek to integrate gender to empower women, need to start with an understanding of the heterogeneity of women (women differ in their socio-economic conditions). The article remarks that care should be taken not to overly focus on a segment of the women category (for example female heads of households) and overstate temporary gender gains that do not transform exploitative gender norms and practices. The article highlights the need for the development of multidimensional programs that can help both married women and female heads of households to gain full access to resources and participate in important household decision-making processes.

Introduction: "Scaling up" and gender integration
A large number of studies highlight that "scaling up" fundamentally relates to replication, i.e., the expansion of successful pilot projects to reach a large number of poor people (International Food Policy Research Institute, 2007;Mangham & Hanson, 2010;Simmons, Fajans, & Ghiron, 2006;UNDP, 2013;Uvin, 1995). Some define "scaling up" as a process of expanding interventions with demonstrated effectiveness "to reach high, sustained, and equitable coverage, at adequate levels of quality" (Victora, Hanson, Bryce, & Vaughan, 2004, p. 1541. Others highlight the impact of external policy contexts, availability and accessibility of resources, delivery methods and pace of change, technological arrangements and evaluation and monitoring strategies on the success of scaling up process (Mangham & Hanson, 2010). Nevertheless, there is no single universally accepted definition and method in use for "scaling up" pilot projects. The approaches and paths developed by funders and implementers to guide the process of "scaling up" have varied (USAID, 2015).
However, scaling up becomes more contentious and complicated when we step away from the issue of replicating successful pilot projects (the quantitative question) and focus on the questions of quality and sustainability of interventions (UNDP, 2013). A range of dynamic, interrelated, and intertwined factors (including social, cultural, political and institutional) affect "scaling up" processes and outcomes (WHO, 2009). In fact, it is not just technologies and best outcomes that are scaled up but also the intricate processes and principles behind the innovations and technologies (CGIAR -NGO Committee & the Global Forum for Agricultural Research, 2001). Hence, adapting service innovation and delivery in the context of changing multi-dimensional factors is a major challenge in "scaling up" success (WHO, 2009).
Often, the success of a "scaling up" process is measured by the degree to which the process is participatory and ensures the inclusion of all stakeholders, especially the poorest segment of society (Binswanger-Mkhize, de Regt, & Spector, 2009). The success of the "scaling up" process is also evaluated by how partners (governmental institutions, funding agencies, NGOs, the private sector, and farmers) understand, view, and respond to the process, and how accountabilities and responsibilities are shared. More importantly, going-to-scale requires organizational and political will and commitment (UNDP, 2013). In addition, success in "scaling up" is not just about expansion and focus on technicalities, it is rather about paying attention to the enjoyment level of social actors involved in the process. This is because without the full participation and support of social actors, efficient and sustainable "scaling up" outcomes cannot be achieved (Uvin, 1995;Uvin, Jain, & Brown, 2000).
Quite commonly, "scaling up" success can also be measured based on the degree to which it addresses gender inequality. Most replicated pilot projects highlight the importance of the integration of gender in the "scaling up" process, a focus that reflects the broader attention given to mainstreaming gender in international development policy. While gender mainstreaming dates back to the 1970s, within the last two decades gender integration in development has been increasingly recognized as critical to bringing about gender equality. The focus on gender integration emerged because even when gender mainstreaming in development programs exists and is strong, gender equality policies do not always connect to the specific intervention objectives and priorities. Gender integration often targets connecting gender equality policies to specific aims and priorities of development programs (SIDA, 2003).
In addition, the increasing attention given to gender integration in development programs, relates to recognition of the roles that women play in economic and social development (Baruah, 2004;Huq & Moyeen, 2011). However, despite the fact that many development programs recognize the need to integrate gender, they differ on how they implement gender integration on the ground, at the grassroots level. Many programs include gender under the banner of "gender integration", partly because donors expect to see gender as an important component of programs that they fund. In other cases, programs recognize that without a focus on women and underprivileged segments of society, poor-friendly programs do not live up to their pro-poor mandate. Several programs consider "gender" as less important priority and they barely mainstream it within their planning, implementation and assessment of programs (Huq & Moyeen, 2011). In terms of differences in approaches to gender integration into programs, USAID's explanatory framework illustrates a range of ways in which gender has been integrated into projects (FHI 360, 2012).
The framework explains the ways in which programs and policies address gender. First, it outlines that some programs and policies do not consider gender important enough to incorporate into project planning and implementation, and hence, are gender blind. However, the majority of projects implemented in developing countries are gender-aware and take measures to accommodate gender in planning, design, implementation, and evaluation. However, these gender-aware projects differ in the manner in which they address gender inequality. Some exploit gender difference to achieve predefined objectives. "For example, a condom ad that portrays men as sexually aggressive or promiscuous exploits harmful norms related to men's sexuality in order to sell condoms" (FHI 360, 2012, p. 9). Many others accommodate gender and work around the existing gender structures and arrangements without actively challenging them. On the other hand, some more progressive projects challenge and seek to transform existing exploitative norms, values and practices and contribute to bringing about gender equality. Accordingly, they fully integrate gender into planning, design, implementation, and monitoring and evaluation of project objectives and activities (FHI 360, 2012).
Meanwhile, despite the prevalence of abundant evaluation resources on gender mainstreaming and integration of gender into health programs, there are limited numbers of studies that are focused on the integration of gender within the context of agricultural programs. This article seeks to contribute to filling this gap by specifically drawing on the example of SPIFoNS, implemented in Southern Ethiopia. The study evaluates the manner in which SPIFoNS integrates gender and the degree to which it has contributed to transforming exploitative gender norms, values and practices. It starts with an introduction of SPIFoNS and description of research methods used. It then describes the strategies used by SPIFoNS to integrate gender and the major gender gains and challenges faced. The aim of the article is not to generalize findings but to highlight some lessons that we may learn from the experience of pulse innovation.

Introduction to pulses and SPIFoNS
The last decade has witnessed increasing attention given to pulse crops within the United Nations to the extent that the year 2016 was named the International Year of Pulses. The same year saw the commencement of the sustainable development goals to end global hunger by the year 2030. The popularity and the recognition of the importance of pulses also emerged in the context of the increase in support for seed aid within and among global development policy makers, such as the World Bank (Rietber, Gevers, & Hospes, 2014). The popularity of pulses is rooted in the benefits that they provide to smallholder farmers in the developing world. Pulses are known to be relatively inexpensive to produce and purchase, and provide plant-based protein, vitamins and minerals. They also allow double cropping and intercropping to happen and contribute to soil fertility and improvement of yields. Moreover, pulses can grow in comparatively poor soils and can also be used as animal fodder (Food & Agriculture Organization of the United Nation, 2016; United Nations, 2016).
Following the recognition of the benefits of pulses for nutrition and food security, pulse based development projects were launched in some African countries such as Ethiopia to expand the production of new high-yielding variety of pulse crops that include lentils, haricot bean and chickpea. The majority of the crop production techniques are aimed at building the capacity of farmers to produce pulses in a sustainable manner through the provision of agricultural and nutrition education and training. Many of them also provide seed credit and seed donation (about 2 kg) for seed multiplication and for the production of grains (Food & Agriculture Organization of the United Nation, 2016; Murrell, 2016).
SPIFoNS emerged in the context of the popularity of pulses and it was implemented in Southern Ethiopia with the aim of contributing to the challenging task of achieving food and nutrition security in the region. The adopted strategy includes using pulse innovations as an entry point to promote systemic transformation from food scarcity and lack towards food and nutritional security of rural households, through scaling-up of intensive pulse crop-based farming systems, agri-food processing and marketing, and nutrition education interventions. SPIFoNS was initiated to catalyze large-scale positive change in food and nutrition security in Southern Ethiopia by scaling up pulse-crop innovations to reach 70,000 farm households. It is funded by the International Development Research Centre (IDRC), Canada, and is jointly led by the University of Saskatchewan, Canada and Hawassa University, Ethiopia. It builds on a 20-year research partnership between the two universities focused on capacity building and training, thus strengthening Ethiopia's capacity to address food and nutrition security and livelihood on smallholder farming communities. Since 2010 the project has focused on pulse crops production; the introduction of high yielding varieties, soil management, agronomic bio-fortification and nutrition interventions to encourage wider use of pulses for household consumption. The project networks, partners, and works with a number of constituencies at international, national, and local levels including the International Crops Research for Semi-Arid Tropics (ICRISAT), Ethiopian Public Health Institute (EPHI), Farm Radio International (FRI), World Food Program (WFP), Southern Regions Regional Health Bureau and Regional Agriculture Bureau of SNNRP, Ethiopia.
The core objectives of the project fall into four interconnected components: agricultural, nutritional, socio-economic processes (gender as a central component) and value chain and marketing. The agricultural dimension of the project targets increasing soil productivity, provision of new improved seed varieties, institution of improved agricultural practices and bio-fortification. The nutritional aspect aims to assist farm households with the introduction of new, and improved, pulse based food preparation and processing. The project seeks to ensure diet diversity and bioavailability and also to assist rural households with value adding and marketing of their produce. The social dynamic and gender aspect of the project seeks to integrate gender in order to empower women. This core objective of the project focuses on making women the beneficiaries of the pulse innovation, through facilitating their full participation in production, consumption, and sales. The focus of this article is on connecting two of the core components of the project, women's empowerment (the challenges of integrating gender) and agriculture.

Research methods
The data for this research emerged from a study undertaken from May to July 2016 and in January 2017, on men and women members of households in the districts of Sodo, Meskan and Halaba, three of the fifteen rural districts in which the project has been implemented. The research took place in the homes and on the farms of the participants.
Observation, in-depth interviews, and focus group discussions were utilized in collecting data. A total of 129 in-depth interviews were undertaken (see Table 1) along with six focus group discussions (two in each of the three districts) entailing three men's and three women's groups respectively. Men in seed multiplication 5 5 5 Single women in grain production 6 6 6 Married women in grain production 9 9 9 Men in grain production 15 15 15 Development Agents (DAs) 3 3 3 Total 43 43 43 The selection of informants for the research was undertaken through utilizing theoretical and socio-economic sampling. The sampling took into consideration a number of socio-economic variables including sex, age, marital, and educational status. Theoretical sampling refers to the process of selecting cases or case groups according to concrete criteria concerning their content instead of using abstract methodological criteria (Flick, 2006). Hence, the research considered a number of criteria such as the number of years that they took seeds from the agricultural bureau, whether they have received haricot beans and/or chickpeas or not, the amount of land they dedicate for growing haricot beans and/or chickpeas, the amount of land they own and family size.

SPIFoNS gender integration strategy
More than 80% of the population of Southern Ethiopia live in rural areas, and women provide the majority of agricultural labor. Women make very crucial socio-economic contribution in Southern Ethiopia. However, there has been persistent inequality between women and men in literacy, access to opportunities, and control over resources and incomes. While they represent the major contributors to the agricultural workforce in general, their contributions are traditionally undervalued and misunderstood. One of the core objectives of the project is to make women the beneficiaries of the pulse innovation through facilitating their participation in production, consumption, and sales. The ultimate objective is to use pulse crops (in particular haricot bean and chickpea) and pulse production technology as vehicles to contribute to the empowerment of women in Southern Ethiopia. Ostensibly, SPIFoNS is a gender-aware project and hence the most important question is evaluating whether the project has just been accommodating gender in its policy and objectives, or if it has actually been transforming exploitative gender norms and practices into sources of empowerment.
In fact, in the process of undertaking the research, it emerged that the project has implemented a number of strategies to integrate gender and to empower women. First, it explicitly planned to make women 50% of the total beneficiaries of its agricultural interventions. The project provides women with agricultural training regarding sowing techniques and the appropriate use of fertilizers and they are encouraged to get involved in the production of the new variety haricot beans and chickpeas. Moreover, women have received training on the nutritional benefits of consuming pulses and, in order to help them increase the consumption of pulses, they attended recipe demonstrations and complementary food-preparation training session. The project has also implemented broader gender strategies including the organization and delivery of gender sensitization workshops and meetings that brought women and men farmers together and openly discussed gender issues. These were aimed at improving men's understanding and recognition of women's roles in agriculture and in the day-to-day household chores.
The research then sought to evaluate the process and the outcome of the gender integration strategy implemented by the project. It considered three main factors as indicators of social transformation and empowerment of women, namely the degree of the participation of women in agricultural activity, the building of the capacity of women, and the extent of the sustainability project outcomes and the gender gains. The study first assessed the degree of women's participation in both seed multiplication and grain production processes. It examined the challenges married women face as participants of the pulse innovation project. It explored how household decision-making patterns are made over production, distribution and consumption. Also it mapped out the nature of the pulse marketing and the place of men and women in it. It highlighted the structure of ownership and control over income generated from sale of pulses. It then outlined the role that the project played in building the capacity of single women. Finally, the research raised the question of the sustainability of the project outcomes.

Participation and empowerment
For more than three decades, the degree of community participation was used as a key criterion in the evaluation of the process and impact of a development intervention (Cummings, 1997;Hickey & Mohan, 2004;Usadolo & Caldwel, 2016). The success of a scaling up process is measured through examining the degree to which the process is participatory and ensures the inclusion of all stakeholders, especially the poorest segment of society (World Bank, 2003). It would not then be incidental in evaluating the degree of success of SPIFoNS in scaling up its programs to ask the extent to which it has been successful in allowing the participation of the poorest segment of society, especially women.
As already outlined, one of the core objectives of SPIFoNS is to empower women and, in its policy statement, the project highlighted the important significance of facilitating the participation of women. In fact, at the time of research implementation, about 41% of its beneficiaries were women. However, the most important question here is what the relatively high percentage of women's participation really informs us about the empowerment of women. Does women's inclusion in the project directly reflect the building of their capacity through their participation in pulse production and consumption?
As stipulated above, the project primarily seeks to empower women through the provision of seeds and agricultural training. Meanwhile, those who are involved in seed multiplication gain more benefits than those who participate in grain production. This is because seed producers get access to more than 20 kg seeds, are provided with fertilizers and pesticides, and are continuously followed up and assisted by Development Agents (DAs). On the other hand, those who are involved in grain production receive only 2 kg seeds, and the follow up and assistance is relatively less. In the process of observation and focus group discussions, it emerged that although all the beneficiaries of the project participate in agricultural training regarding sowing techniques and the use of fertilizers, the degree of attention given to grain producers compared to those who produce seeds is considerably less. Also, those who are involved in seed multiplication are relatively well off, have about a quarter of a hectare of land, and are networked because seed is produced by farmers who are organized in clusters (groups) and have adjoining plots of lands. Hence, the most important question is to what extent women participate in seed multiplication, which is relatively speaking more empowering?
Contrary to the aggregate women participation of 41%, the number of women who participate in seed multiplication in the three research sites is small. Only 30% of seed producers are women and all are single (Henry & Beyene, 2017). It is evident that apart from married women, the project targets single women and primarily provides its services to them more than to married women. Meanwhile, despite their lack of participation in seed multiplication, a significant number of married women participate in grain production. More than 45% of women who are involved in grain production are married women. One salient method of evaluating the degree of the empowerment of women is to examine the degree of their participation in pulse production, consumption, and trading decision-making. In fact, many critical development studies literature and gender integration frameworks establish direct links between women's empowerment and control over household decision-making (USAID, 2014). Hence, the fundamental question the research asked was that, further to providing seeds and agricultural training, to what extent the project facilitated the participation of married women in important household decision-making on pulse production, trading and consumption. Examination of the household decision-making process on production, distribution and consumption helps us to understand whether the project has just been accomodating gender or has contributed to transforming existing gender structures and exploitative gender norms and practices.

Gender integration: Production
In order to assess the contribution of SPIFoNS to the empowerment of married women, the study first examined the degree of married women's participation in production decision-making. It sought to identify the extent to which women participate in important production decision-making regarding what to produce, how much to produce and when to produce. The study assessed the responses of women and men on production decision. As illustrated in Graph 1, out of the total interviewed women only 5% of them said that they make production decisions on their own. While about 90% said they do not make decisions, about 5% said they make decisions in consultation with their husbands.
When men were interviewed regarding decision-making over production, a similar pattern emerged with a slight variation. While about 60% of them said they make decision on their own, another 30% asserted that they make decisions in consultation with their wives, whereas 10% of them said their wives have the final say.
The majority of the interviewed men claimed that they make decisions on what to produce because they believe that women do not know about agriculture and do not plough land. Men argued that they decide on what to produce because they undertake the hard task of ploughing the land. They rationalize their power over agricultural decision-making on the basis of ploughing the land. On the other hand, interviewed women outlined that, although in theory they have control over the land registered under their name, in practice, in all the three research sites their power to exercise their rights over land and property was very limited. Despite their right to inheritance, when they get married, women often leave their land to their brothers or male relatives. Also, after they get married, in theory women get equal rights over land inherited or owned by their husbands. However, in practice, they have limited direct power to make decisions over the use and management of land. According to custom and practice, husbands and male family members and relatives have culturally made decisions regarding land use.

Gender integration: Time poverty
However, despite men's assertion that women do not know about ploughing and agriculture, women are over-burdened by agricultural and domestic work, which is often not recognized and respected. In fact, as a result of being subjected to exploitative cultural relations, the majority of the women participants of the research not only suffer from extreme resource poverty and deprivation, but also are victims of time poverty. In the three research sites women were found to spend more than 15 h per day on agricultural and domestic work, whereas men spent only about 8 h on agricultural and domestic work (illustrated in Graph 2). Men spend about one third of their time on socialization, whereas women have very limited time for socialization and for rest.
When we disaggregate the data based on considering the time men and women spend on leisure and socialization in each of the research sites, it emerged that time poverty extremely affects women in Halaba, more than Sodo and Meskan. In Halaba while women spend less than an hour on socialization and leisure, men spend about 10 h per day socializing (see Graph 3). In the studied sites, in Halaba in particular, it is men who participate in important meetings such as kebele (local government) meetings, whereas women's mobility is restricted and they often spend their time engaging in domestic and agricultural activities.
Compared to men, women have very little time to go out and establish networks and links with other women. Married women's access to information is predominantly through their husbands. In the three research sites, men had greater mobility and established networks with local community

CROP PRODUCTION
leaders, local government officers, agriculture and health extension workers, and crop retailers and buyers. While men's mobility and their networks give them power, women's limitation within the domestic domain is a major cause of their disempowerment and marginalization.
Often women are busy undertaking heavy household work as well as working in the fields. Women wake up early in the morning and go to bed late at night. While men can normally wake up late in the morning and go to bed early, women start the day by gathering the dung of the livestock. Then they fetch water (in a plastic tank or a big clay pot), carrying it on their backs, which can take between a half hour to two hours depending on their location. They prepare food for the family. In the absence of older siblings, women undertake child minding. Although women do not plough land, they undertake the planting, weeding, tending, and harvesting of crops. Women are also responsible Graph 2. Time poverty in the three research sites.

Graph 3. Time dedicated to leisure and socialization in the three research sites.
for feeding and raising cattle and minding chicken. Men primarily undertake ploughing. However, while men have some leisure time during which they visit the local pubs, women are mostly busy with their domestic tasks.
While women represent the major contributors to the agricultural workforce in Southern Ethiopia in general, their contribution is undervalued and misunderstood. Men's roles in agricultural activities are clearly established, while women's roles are not as clearly defined and recognized. Despite women's active and direct involvement in agricultural activities, men often do not assist women with the household activities. Some men, in Halaba in particular, spend their day chewing chat (a type of mind stimulant leaf that is classified as a drug), whereas women undertake most of the household activity and the minding of children as well as cattle. Hence, it is evident that the nature and structure of property ownership and the division of labor disadvantage women. The provision of agricultural education and pulse seeds to married women farmers plays a very limited role in the transformation of these exploitative structures and practices.

Integration of gender: Consumption and sale of produce
Another area that the research evaluated was the degree of success of SPIFoNS in enabling men and women to make joint decisions over the consumption and sale of pulses produced. As illustrated in Graph 1 compared to production, women's participation in decision-making over consumption and sale of pulse products is much higher. More than half of the men interviewed, and slightly less than half of the married women interviewed, noted that they make joint decisions over what to consume and sell. With respect to decision-making on consumption, three interviewed married women noted that before their husbands take pulse products to the market, they consult them because often husbands do not know how much of the produce needs to be kept for home consumption. Because married women do the cooking, husbands often consult their wives regarding what needs to be kept for home consumption. However, with respect to sale of produce, men still dominate decision-making and married women often do not have as equal say as men on when and how much of the produce to sale.

Integration of gender: Differential access to markets
Another key area that the research looked at is women's access to and participation in pulse markets. In this regards, observation of some of the markets in the three research sites revealed that markets are noticeably dichotomized into men's and women's domains. While the women's section of the market often provides consumables and some relatively cheaper products such as eggs, cheese, butter, chat, pepper, and utensils such as Akenbalo (baking lids), the lucrative market for farm animals and fattened cattle is under men's control. With respect to purchases, while women are in charge of going to market to buy household necessities such as soap, vegetable oil, and table salt, men undertake the buying of farm animals, cattle, and farming tools.
With regard to sale of pulses, when small quantities of pulses such as chickpea and haricot bean (usually less than 20 kg) are to for sale, women take it to the nearest market, carrying it in a basket on their backs. However, when larger quantities are to up for sale (quintals or at least 50 kg), men take it to the larger markets packed in big sacks carried by donkeys. While women retail small quantities of pulses, using a jug or a tin as a measure of unit, large quantities are measured by weighing scales. Men often claim that they do not know how to sell small quantities using a jug (tin) as a measuring unit. Even quantities of up to 20 kg are considered too small for men to take to market because when men go to market they are expected to go to local pubs and buy drinks for their friends. Accordingly, while men control the higher proportion of the household income, women control small incomes.
More than 70% of men argued that they take large quantities of crops to the markets because women do not know about weighing scales and can be easily cheated. Also, the majority of men say women cannot balance sacks on the backs of donkeys to take produce to distant markets. Some of the men claimed that their wives did not know about money and could not count. Others think women are not able to take care of money, and that they waste money. However, the majority of women asserted that they do know about the weighing scales, can count money, and can easily balance sacks on the backs of donkeys, and that what men say in this regard is untrue.
Women outlined that while most men get drunk when they go to market, the women do not drink, and hence are less likely to be cheated on the weighing scales. Also, they do not waste money on drinks. The main reason why women do not take large quantities to markets is because their husbands do not want them to control the finances. Some men do not even allow their wives to take small quantities of crops to markets, and hence they steal (hide from their husbands) when they take the small amounts to the market. Even if it is also their property, women on occasions hide from their husbands when they take crops to markets. Meanwhile, 50% of interviewed married women also noted that often, social expectations and domestic workloads compel them to go only to the nearby small markets to sell small quantities of crops.

Gender and control of income
One of the important questions to ask in evaluating the success of SPIFoNS in the integration of gender in its program is to what extent it has been helping women to gain control of income? With respect to this, when men and women were interviewed, significant differences emerged. While 65% of interviewed men said they and their wives have equal control over income, 85% of women interviewed said that men made decisions on their own (see Chart 1).
On the other hand, 5% of both men and women interviewed said that women control income. What appears to be evident is that men dominate control over income generated, and they control the majority of income from the sale of pulses in the three research sites.
On the other hand, more than 70% of women outlined that they control the small money that they earn from the sale of crops (which they take in baskets), and also control the income they generate from running small enterprises located in their homes. In Sodo in particular, women are actively involved in running their own income-generating businesses, and they do not expect to receive money for household expenses from their husbands. Six of the fifteen women individually interviewed in Sodo are engaged in small-scale home-based activities such as the making of Areke (local alcoholic drink) and Enjera (local bread), and retailing chat and pepper. However, as the business activities in which women are engaged are small-scale, the amount of income they earn is small. Nevertheless, women control the income they generate from these activities and use it for household consumption.
While most men use the majority of the income to buy agricultural inputs (fertilizer, seeds, pesticides, and herbicides), paying tax, and paying for labor used in harvesting crops, they nevertheless retain some money for their own personal needs. According to the interviewed married women, some men use the money for alcohol and Chat (local mind stimulant classified as drug) and become unable to buy fertilizers and pay taxes. To cover up their bad deeds, some of them become arrogant and start fighting with their wives. One woman who was tired of her husband's behavior in particular noted that she does not like the days her husband goes to the market because he often comes back drunk and starts fighting and beating her.

Capacity building and sustainability
In the process of undertaking the research, it emerged that more than supporting married women, the main achievement of SPIFoNS is in assisting the building of the capacity of single women (divorced, separated and widowed) and married men. The majority of interviewed single women and married men outlined a multitude of advantages they gained including increase in income and assets and in the reduction of their vulnerability. For example, as illustrated in Graph 5 more than 60% of interviewed women that are involved in chickpea seed multiplication in Sodo and Meskan noted that they produced more than 500 kg of chickpea per 0.25 hectare and sold 25 birr (1.3 CAD) per kilogram. Similarly, the majority of farmers who participated in haricot bean seed multiplication in Meskan produced more than 600 kg. of haricot bean per 0.25 hectare and sold it for 16 birr (about 1 CAD) per kilogram (see Graph 4). Also, women farmers used about a third of their produce for consumption and prepared a range of food items, thus enabling them and their families to be healthy and strong.
Moreover, involvement in production of pulses has enabled women to expand their asset base. For example, five women heads of households bought oxen and no longer rely on sharecropping. Another three female famers maintained that they bought radios, thereby gaining access to agricultural information. Some women were able to buy clothes for their children and are now able to send them to school. These observations are significant indicators of integration of gender that has enabled single women to become involved in producing pulses, a development which has assisted them to become economically empowered.
In addition, involvement in pulse production has contributed to the reduction of women's vulnerability and the enhancement of their social mobility. With regard to this, while some of the interviewed single women said that they have saved money for an emergency, others said that they no longer borrow from moneylenders and relatively well-off relatives. More importantly, the majority of interviewed single women heads of households said that their role in society has increased in value, that they felt important and got respect from others. This in particular, relates to the agricultural training that they received. Interviewed women appreciated the lessons regarding sowing techniques and the use of fertilizers, and they now feel like experts, sharing their knowledge with others who did not attend.
In connection with capacity building, the other important question the research evaluated was the sustainability of the project achievements. Sustainability in this context relates to the continuation of the development project activities and maintenance of project outcomes after exit (Joshi, 2016). The research sought to identify evidence, which indicates that the project's gender gains will be sustainable. In the process of undertaking the interviews and observing what women and men do, two main indicators of sustainability became apparent.
In the three research sites, almost all the women who participated in the research saved some of their produce as seed, and some also shared seeds with their relatives and neighbours. As can be seen on Graphs 4 and 5, those involved in haricot bean seed multiplication saved from 20 to 30 kilos of their produce, and those who are involved in chickpea seed multiplication saved between 30 and 35 kilos of their produce.
Similarly those women who are involved in haricot bean (grain) production save 10-20 kilos of their produce, while those who are involved in chickpea (grain) production save from 8 to 10 kilos of their produce (see Graphs 6 and 7).
When women were asked the question, "will you grow pulses even if you were not provided with seeds, fertilizer and pesticides?", the majority responded positively. This culture of saving and sharing seed indicates sustainability of project outcomes, and its success in contributing to the expansion of women's livelihoods in a sustainable way.
Another important indicator of the sustainability of the outcome of the project is the increase in the sizes of land dedicated to pulse production. In the three research sites, there was a significant increase in the size of land dedicated to the production of pulses. A comparison of the sizes of land dedicated to the production of pulse two years before, and during the winter of 2017, revealed that there has been a significant increase in the size of land dedicated to pulse production in the three research sites (see Table 2). More than half of the interviewed women provided two reasons for the increase in the size of land they allocated for the production of pulses.
First, they asserted that the new variety of pulse seeds introduced by the project provided more yields than the old variety that they used to grow. The women also noted that the training that the project provided in new agricultural methods (sowing and use of fertilizers) improved their agricultural knowledge and the application of the new knowledge contributed to increased yields. In a nutshell, the project's focus on explicitly assisting single women to become involved in producing pulses contributed to the creation of sustainable agricultural (pulse production) systems that has the potential to empower single women in a sustainable manner.

Conclusion
This article critically illustrated the methods adopted by SPIFoNS to integrate gender and to empower women. Despite its fundamental mandate of integrating gender in the scaling up process, SPIFoNS seems to have mainly focused on providing its services to single women. Notwithstanding the project's objective of empowering women, the number of women in the relatively lucrative practice of seed multiplication is very small. Overall, men dominate seed multiplication and women who are involved in seed multiplication are single (widow, separated and divorced).
On the other hand, the article demonstrated that married women who participate in grain production have limited power over production, consumption, sale of produce and income generated. Similarly, women's participation in pulse markets is minimal and is focused on the lower and less profitable segment of the market whereby small quantities of pulses are traded. In addition, the research demonstrated that while married women play a key role in agriculture, including pulse production, their contribution is generally understated. Married women in the three research sites are not just victims of resource poverty, but also time poverty.
The main achievement of the project in terms of empowering women is that it has contributed to the building of the capacity of single women by helping them to earn additional income and increase their assets, which in turn results in the reduction of their vulnerability. However, despite the achievement of the project in terms of assisting single women, it is crucial to highlight that the integration of gender and the empowerment of women is not just about contributing to empowerment of single women. Paying attention merely to single women, without assisting married women that face huge gender barriers in the household, is accommodating rather than contributing to the transformation of the existing exploitative gender structures. Hence, rather than accommodating the existing patriarchal gender structures through mainly focusing on single women and overstating temporary gender gains, SPIFoNS and other similar projects need to design broader strategies that contribute to the the transformation of exploitative gender norms and practices located in the domestic domain.  They need to contribute towards changing patriarchal gender structures that restrict married women from fully participating in important decision-making processes in the areas of production, distribution, consumption and trading of agricultural crops such as pulses.

HUSBAND WIFE HUSBAND AND WIFE
In addition, SPIFoNS project and other similar projects that are implmented in Sub-Sharan Africa, need to incorporate policy structures and practical guides that facilitate the recognition and validation of women's contribution to agriculture. They need to find ways to facilitate the participation of married women in lucrative productive activities such as pulse seed multiplication. Moreover, rather than one-size-fits all kind of approach in dealing with women, they need to be sensitive to the heterogeneity of the category -women and pay attention to the differences in their socio-economic conditions and social positions. While accommodating gender may help to highlight positive outcomes and short-term gender gains to satisfy funders, such an approach in the long-term will not contribute to reducing culturally entrenched gender hierarchy and resultant inequality.
One crucial way of brining about long-term positive change is enabling married women to build their social capital. This can be achieved through helping them to be organized, and form association, links and groups. Helping married women in establishing women's cooperatives, for example, for the production, consumption and processing, and selling of pulses would help them to be empowered. They would mobilize their resources and work together to improve their socio-economic circumstance. The formation of cooperatives would help create a forum for married women to share information, knowledge and experience and these ultimately help them to empower themselves.
Another important way of transforming the deeply seated exploitative cultural values and practices is the implementation of sustained participatory gender analysis and gender sensitization programs. The creation of a forum whereby men and women reflect on gender relation and their experience, would assist men to better understand the exploitative cultural values and practices that disadvantage women. In addition, the empowerment of married women necessitates the joint efforts of governmental, nongovernmental, and local and international development organization. Hence, improving the contribution of SPIFoNS project to empower married women demands the project's ability to strengthen the existing and to create new networks with NGOs and other cooperatives. The creation of links with other cooperatives such as credit and savings cooperatives would enable married women to get access to not just pulse seed credit and agricultural training, but also finance to expand their agricultural activities.
Finally, the authors would like to note that the extent of the success of SPIFoNS project to empower women has been affected by complex factors including politics, institutional process and governance, agency and legitimacy. This article has focused on elucidating the socio-economic and cultural factors, and hence constructing a comprehensive analysis requires examining other crucial factors. Moreover, a more theoretically informed analysis of the empowerment process demands relating SPIFoNS project's objectives with the broader and controversial development themes such as development aid. Hence, future research on the topic may focus on filling the theoretical and empirical gap created by this article.