Using Agriculture Extension Agents to Promote Nutrition: A Process Review of Three Feed the Future Activities in Ethiopia

Preparing for a cooking demonstration. Photo Credit: T. Yadetie, for USAID ENGINE

In March and April 2014, SPRING conducted a rapid, participatory process review to document experience and learning from activities that are utilizing a common strategy to promote nutrition through Feed the Future agriculture investments in Ethiopia. This strategy involves training agriculture development agents (DAs) to deliver nutrition social and behavior change (SBC) messages and interventions to support the integration of nutrition and agriculture. This review examined the experience of three Feed the Future activities: Empowering New Generations to Improve Nutrition and Economic Opportunities (ENGINE), Agricultural Growth Program – Agribusiness and Market Development Project (AGP-AMDe), and Pastoralist Resiliency Improvement and Market Expansion (PRIME). Capturing the process in Ethiopia contributes to the Feed the Future learning agenda and may also help refine agriculture-nutrition approaches in current and future activities around the world.

The objectives of the review were to:

1. Document the coordination among Feed the Future partners that are utilizing agriculture extension workers to deliver nutrition SBC messages and interventions; and
2. Capture the process, challenges, successes, and lessons learned from the perspective of project staff, government counterparts, agriculture extension workers, and community members participating in the activities.

The SPRING team reviewed relevant documents, conducted key informant interviews, and facilitated focus group discussions with relevant stakeholders in Addis Ababa, Afar region and Amhara region to document the experience of the three Feed the Future investments. The review team met with activity staff, government counterparts, DAs, and community members at the national, regional, woreda, and kebele levels.

This report outlines challenges and successes in coordination within and between the three Feed the Future activities and between the activities and the government. It also documents the process, challenges, successes, lessons learned, and gender aspects associated with use of agriculture extension workers to deliver nutrition messages. The main findings describe the implications of differing activity designs and targeting approaches, the importance of context, and the challenges of working within the larger DA system.

This review was not intended to assess the effectiveness of this approach, but to inform further exploration of its relative merit compared to other SBC interventions, and what constitutes better practices for designing, implementing, and monitoring this approach. The insights garnered from activity implementation experience and different stakeholder perspectives resulted in several key recommendations to enhance the nutrition sensitivity of current agriculture interventions and to design more nutrition-sensitive agriculture investments in the future. They are summarized below:

• Context should inform initial activity design as well as mid-course corrections.
• Challenges within the DA system can be addressed by building on recent political attention and prioritization being given to nutrition in Ethiopia. Nutrition-sensitivity of activities can be enhanced by applying the agriculture-to-nutrition conceptual pathways during annual work-planning and/or design of mid-course corrections.
• Per the National Nutrition Program, mechanisms to support cross-sectoral collaboration and increase leveraging of resources should be strengthened at national, regional, zonal, and woreda levels.
• ENGINE has an important role to play in continuing to work with the USAID | Ethiopia Mission on developing comprehensive social and behavior change guidance.
• Not all agricultural interventions should target nutrition outcomes at the household level.

Introduction

Objectives of the Process Review

A cooking demonstration. Photo credit: T. Yadetie, for USAID ENGINE

In March and April 2014, the Strengthening Partnerships, Results, and Innovations in Nutrition Globally (SPRING) project conducted a rapid, participatory process review to document experience and learning from three different Feed the Future activities that are utilizing a common strategy for promoting nutrition in Ethiopia. This strategy utilizes agriculture extension agents to promote maternal and child nutrition by conducting social and behavior change (SBC) interventions, including nutrition messaging. The activities included in this review include: Empowering New Generations to Improve Nutrition and Economic Opportunities (ENGINE); Agricultural Growth Program – Agribusiness and Market Development Project (AGP-AMDe); and Pastoralist Resiliency Improvement and Market Expansion (PRIME). The objectives of the review were to:

1. Document the coordination among Feed the Future partners that are utilizing agriculture extension workers to deliver nutrition SBC messages and interventions.
2. Document the process, challenges, successes, and lessons learned from the perspective of project staff, government counterparts, agriculture extension workers, and community members participating in the activities.

This strategy is being used in several Feed the Future target countries¹ to increase and improve nutrition outcomes from agricultural, health, and nutrition activities. However, the way in which agriculture extension workers are being engaged to deliver SBC interventions for nutrition differs across activities and countries. This review attempts to better understand similarities, differences, and rationales associated with the use of this programming approach. This report should be useful to Feed the Future stakeholders in Ethiopia in the following ways:

• To support the Mission and implementing partners in their adjustments to current programming and coordination.
• To inform the design of a next round of nutrition-sensitive agriculture activities.
• To inform the ongoing process in which ENGINE and other Feed the Future activities utilize a common approach to SBC for nutrition in their activities.

Ethiopia is an ideal context to carry out such a process review for the following reasons:

• Program learning is a priority for the Mission and implementing partners.
• There are several Feed the Future activities that have been implementing for a year or more.
• The ENGINE activity is in place to provide technical assistance for maternal and child nutrition during the critical first 1,000 days, reaching households living in the Feed the Future zone of influence. ENGINE also plays a critical role in coordinating nutrition interventions among implementing partners that have activity objectives related to nutrition.

Preliminary discussions with ENGINE established that there was also interest in using this review to inform a planned internal midterm assessment. Specifically, ENGINE and the Mission asked SPRING to explore:

• How ENGINE has collaborated and provided guidance to Feed the Future partners in cross- training agriculture extension workers and how this can be improved or enhanced.
• How Feed the Future partners, including ENGINE, have been coordinating their community-level activities utilizing agriculture extension agents to deliver nutrition SBC and how to improve this coordination.

Background

Farm household. Photo credit: T. Yadetie, for USAID ENGINE

As Feed the Future activities grapple with how to improve nutrition outcomes among the vulnerable populations in their zones of influence, many of them have chosen, seemingly independently, to implement a similar strategy or approach. This strategy strives to integrate the delivery of SBC interventions that promote nutrition into the roles of existing community-level agriculture workers and volunteers.^2,3 The process, time, and resources involved in integrating nutrition SBC into activities using this approach are poorly understood. Fanzo et al. (2013) recently provided a synthesis of experiences and lessons in integrating nutrition into extension and advisory services covering case studies from 12 countries, including Ethiopia.⁴ The conclusions, however, are broad, and the recommendations are not specific to each country analyzed. The case study on Ethiopia provided an overview of the efforts of the Government of Ethiopia but not the integration of agriculture extension and advisory services with the health extension program. Fanzo et al. further note the importance of behavior change and the need for additional study to better understand how extension and advisory services can play a role in providing the knowledge and skills needed for dietary behavior change.

SPRING’s “Leveraging Agriculture for Nutritional Impact through the Feed the Future Initiative: A Landscape Analysis of Activities Across 19 Focus Countries” identified commonalities of the approach across countries, including:

• Adapting parts of existing maternal and infant and young child nutrition training and counseling packages, sometimes including water, sanitation, and hygiene (WASH) materials for use by agriculture extensionists.
• Implementing context analyses to tailor messages for use by agriculture extensionists to increase target groups’ uptake of promoted behaviors.

Most Feed the Future activities in Ethiopia, including ENGINE, AGP-AMDe, and PRIME, are employing a range of different strategies for utilizing development agents (DAs) to promote nutrition through SBC interventions. Each activity is following a slightly different process based on a) its stage in the activity life cycle; b) whether the nutrition SBC interventions were part of the initial activity design; and c) its differing roles and objectives within the Feed the Future portfolio. Because implementation is still in early stages, neither the planning, implementing, and monitoring of activities nor the successes, challenges, and lessons learned to date have been systematically documented. This report is a step toward addressing that gap.

Activity Description

Selection of activities for this review was based on a combination of recommendations from the Mission and interest on the part of the focal activity leadership and staff. A brief description of each of the three activities reviewed follows.

ENGINE is a large- scale ($50.8 million) multisectoral nutrition activity that aims to improve the nutritional status of Ethiopian women and children under the age of five through sustainable, comprehensive, coordinated, and evidence-based interventions. The major components of the activity, as described by its intermediate results (IR), include building capacity and coordination mechanisms within the Government of Ethiopia to strengthen nutrition programs and policy (IR 1); improving the quality and delivery of nutrition and health care services (IR2); preventing undernutrition through community‐based nutrition care and practices (IR3); and adopting a rigorous and innovative learning agenda (IR4). ENGINE targets vulnerable households and implements health- and nutrition-specific activities through the Ministry of Health (MOH) and health extension workers (HEWs), as well as livelihood and nutrition-sensitive activities through the Ministry of Agriculture’s (MOA) DAs. Nutrition-specific activities account for 80 to 90 percent of total activity resources.

Under IR3, ENGINE provides vulnerable households with livestock and crop (fruit and vegetable) inputs and technical assistance through DAs. The inputs provision is expected to increase the amount of fruits, vegetables, and animal products available to vulnerable mothers and children for consumption of a more diverse and nutritious diet. It is also expected to provide supplementary income to the household so families can purchase other nutritious foods. In addition, technical assistance provided by DAs aims to raise nutrition awareness, increasing the demand for diverse diets, nutrition services, and hygiene among the target vulnerable households. These, together with programmatic inputs from the other IRs, contribute to the activity’s ultimate outcomes, which are (i) improved infant and young child feeding and maternal nutrition practices, and (ii) increased household consumption of diversified diets.

At the community level, ENGINE trains DAs in nutrition-sensitive agriculture and nutrition-specific messaging, which is focused on diversifying agricultural production. ENGINE has also developed a training manual for DAs, as well as a set of resources for health and agriculture program managers. ENGINE zonal coordinators provide supportive supervision to vulnerable household beneficiaries. They conduct spot checks at the farmer training centers (FTCs) and schools, where trainings and cooking demonstrations are conducted and demonstration gardens are installed. Meanwhile, ENGINE regional livelihood coordinators provide supportive supervision to woreda agriculture officers, focusing on the integration of ENGINE activities into regular woreda agricultural extension support. The nutrition technical support that ENGINE provides is the first time maternal and child health and nutrition have crossed over into the mainstream work of DAs and woreda agriculture officers. Human health and nutrition were previously handled exclusively at the grassroots level by HEWs.

AGP-AMDe (Agricultural Growth Program – Agribusiness and Market Development Project) is a $50 million activity to support the Government of Ethiopia’s Agricultural Growth Program. AGP-AMDe is designed to strengthen select agricultural value chains, including maize, wheat, coffee, sesame, honey, and chickpea, in high-growth potential woredas within the productive zones of Ethiopia. It aims to stimulate innovation and private sector investment, increase access to financial services, and strengthen competitiveness among smallholder farmers. Activities are organized around four IRs: improved competitiveness of select value chains (IR1); improved access to finance (IR2); improved enabling environment of the select value chains (IR3); and increased innovation and investment (IR4). Nutrition is identified as a cross-cutting theme, as are gender, behavior change communication (BCC), and knowledge management and information and communication technology.

Nutrition, along with gender and BCC, is integrated into the value chains through ACDI/VOCA’s trademarked Sell More For MoreTM (SMFM) training programs. AGP-AMDe uses a cascade training strategy beginning with a training of trainers with MOA home economics agents (HEAs). The HEAs then teach the nutrition-sensitive SMFM modules to DAs, who then carry out the same training with farmer cooperatives whose members are the target beneficiary clients of the activity. Each HEA trains 20 DAs at woreda level, and each DA then trains 40 farmers within the cooperatives they work with. There are very few female members, so the project is asking DAs to train 20 farmer members along with their wives while also trying to increase the number of women cooperative members. Nutrition-sensitive agriculture training under SMFM aims to improve overall understanding of diet diversity, the importance of intercropping, year-round production for home consumption, the role of micronutrients, hygiene, and budgeting to buy nutritious foods year round.

PRIME (Pastoralist Resiliency Improvement and Market Expansion) is a $52 million activity that focuses on pastoralist households to improve their livestock production and marketing systems (IR1); enhance their resilience and adaptive capacity to climate change (IR2); increase and diversify the household assets of the chronically food insecure and vulnerable, especially those transitioning out of pastoralism (IR3); improve nutritional outcomes (IR4); and improve knowledge management and learning for pastoralist issues (IR5).

Under its nutrition objective, PRIME aims to increase the demand for nutritious foods by leveraging activities from its other IRs, such as increasing livestock productivity. For instance, by increasing access to animal fodder year round, mothers and children can gain access to animal milk and other animal products, which would improve nutritional outcomes for children. PRIME intends to create demand for diversified foods and establish a strong social and behavior change communication (SBCC) strategy to address malnutrition, including edutainment programs for target households and communities involving radio, comics, posters, and live performance. In addition, PRIME trains DAs on the importance of diet diversity and HEWs on Essential Nutrition Actions.

Methods

The participatory process review comprised a secondary document review, key informant interviews, and focus group discussions as described below.

Review of Activity Documents

The review team examined context assessments, annual reports, work plans, activity monitoring plans, behavior change strategies, training materials and job aids, and supervisory support and reporting forms. This document review, along with planning discussions with activity managers and the Mission, informed the development of data collection instruments and selection of activity sites and respondents.

Primary data collection consisted of key informant interviews with key Feed the Future activity staff, government counterparts at the regional and woreda levels, and DAs, community-level agriculture extension agents, as well as focus group discussions with activity beneficiaries. Interviews and discussions explored each stakeholder’s perceptions of the process, successes, challenges, and lessons learned related to delivery of nutrition SBC by DAs. SPRING led small collection teams in each region consisting of local language interpreters who were experienced focus group facilitators. Activity staff helped local staff and government offices coordinate site visits. Staff also participated in several government interviews as etiquette required but did not sit in on focus group discussions or DA interviews. ENGINE staff reviewed draft data collection instruments ahead of data collection to provide an opportunity to add questions relevant to its upcoming internal midterm assessment.

Key Informant Interviews and Focus Group Discussions

Focus group discussion in Amhara Region. Photo credit: SPRING

The review team conducted key informant interviews with activity staff based in Washington, Addis Ababa, and regional offices in Amhara (for AGP-AMDe and ENGINE) and Afar (for PRIME). In addition, regional and woreda-level government officials and DAs were interviewed. For a complete list of names and titles of those interviewed, see Annex 1.

SPRING facilitated focus group discussions with men and women in the community working with PRIME in Amibara woreda, Afar; discussions were facilitated with men and women working with ENGINE in South Achefer and Bore woredas, Amhara. AGP-AMDe was still in the process of training DAs, who had not yet begun promoting nutrition with farmers, so no community-level discussions were conducted.

SPRING developed a focus group discussion guide and separate key informant interview questionnaires for project staff, government counterparts, and DAs. ENGINE reviewed draft data collection instruments and provided feedback to ensure that the questions explored were of interest and would be helpful for the upcoming internal assessment. The instruments are attached in Annex 2. A local research firm translated the instruments into Afari and Amharic and provided experienced focus group facilitators to support the data collection. Key informant questionnaires explored activity background, process, coordination with government and other Feed the Future activities, challenges, successes, and lessons learned. The focus group discussion guide explored community perceptions of hunger, malnutrition, food insecurity, DAs and other nutrition-related actors in the community, access to DAs and exposure to nutrition messages and interventions, and factors relating to uptake of nutrition practices being promoted by DAs.

Data Processing and Analysis

Multiple team members took detailed written notes during the interviews and focus group discussions. The data collection team processed all of the notes at the end of each day and entered them as a group into the questionnaire or discussion guide format. The research team developed an analysis matrix around the areas explored (process, coordination, challenges, successes, and lessons learned) and any emerging themes. The only additional theme that emerged from the data was gender. The team leads then processed the field notes into the analysis matrix by project and type of respondent.

Strengths and Limitations of Methods and Issues Encountered During the Survey

The qualitative methods used allowed respondents to guide the discussion and resulted in deeper exploration of different perspectives. Triangulation among different respondents and project documents validated the findings and allowed some generalization within and across projects using similar strategies.

Weaknesses resulting from the limited number of project sites visited and respondents—particularly of DAs who had been implementing for some time—included an increased chance of overstating or understating any given theme or issue or missing an issue altogether.

Because focus group discussions and some key informant interviews were conducted in local languages the quality of SPRING’s translators was very important. While the principal investigators reviewed findings closely with the translators each evening, it is important to note that there is always a possibility that some nuances will be lost in the final analysis any time translation is involved in data collection.

Discussion of Findings

A Health Extension Worker leads the demonstration. Photo credit: T Yadetie for USAID ENGINE

The findings are detailed below according to the main categories of questioning used in the survey: coordination, process, challenges, successes, lessons learned, and gender. The review team anticipated different perspectives among different stakeholders and conducted the interviews in an open manner to facilitate a diversity of perspectives. Analysis of the data indicates that the views among project staff (whether centrally- or field-based), government counterparts at the regional and woreda levels, and DAs on coordination and various aspects of the process are generally very similar to each other. It is notable, though, that the responses of those stakeholders differed from those of both male and female community members. The text below indicates to which group of respondents the findings correspond wherever possible. Finally, findings related to ENGINE’s collaboration with and support of other Feed the Future implementers are included throughout, as they relate to the topic areas presented.

Coordination

This review prioritized the examination of coordination, as it was a key objective of the exercise. The USAID | Ethiopia Mission also believes that efforts at coordination are important in obtaining cost-effective and sustainable outcomes. Therefore, the Mission and ENGINE were interested in learning whether and how ENGINE was supporting coordination among and across the other Feed the Future activities to improve the coverage and quality of nutrition interventions. Specifically, the review examined coordination between the target activities and government entities, between and among Feed the Future activities, between DAs and HEWs, and between Feed the Future activities and other community-based entities.

Coordination with Government

At the national level, a significant challenge for coordination with the government is that the national- level MOA has not fully embraced the national nutrition mandate, despite being a signatory to the National Nutrition Plan. ENGINE and others continue to do advocacy with the MOA at the national level. They also have engaged parliamentarians to facilitate their involvement with nutrition and to have them act on it as an urgent priority. An AGP-AMDe manager suggested that one way to improve multisectoral collaboration within and between government and nongovernment partners would be to strengthen the National Nutrition Forum. The forum is a multisectoral government body, but it is relatively inactive.

All three activities coordinate with government counterparts at the regional, zonal, and woreda levels around utilizing DAs to promote nutrition. However, ENGINE is the only activity that seconds staff in MOA offices at the zonal and woreda levels. Key challenges involved with this level of collaboration include 1) the wide geographic distribution of activity staff, making supervision difficult; and 2) zonal and woreda offices’ preference that ENGINE pool activity funds with other government budget sources. A manager with ENGINE summed up the challenges and rewards of collaboration:

At the regional and zonal level, our people sit in the government office of their respective sector, health with health, ag with ag. That makes things move the best. Initially we couldn’t get space for these zonal coordinators to sit. They [the government offices] were thinking, it is an NGO, it is your business, go away. Then they were convinced that we are supporting them to do their own job. They see what is happening and they see their own improved performance. I think we are now good.

ENGINE and AGP-AMDe often work with the same government counterpart at the regional and zonal levels. One such government respondent, a zonal agriculture extension head, felt that coordination among all three actors was going well. He attributed this in part to the fact that each activity has different targets and objectives.

I coordinate those working at woreda level. AMDe includes all farmers, while ENGINE has targeted beneficiaries who are vulnerable. AMDe is one arm of AGP. The goal of AMDe is highly related on business. AMDe does training only, while ENGINE provides inputs. I am the contact person for both projects, and ENGINE and AMDe are working together.

At the regional and zonal levels, some government stakeholders within the agriculture sector were more enthusiastic than others, depending on whether they had a budget for nutrition. One project manager with AGP-AMDe commented:

Nine ministries signed the National Nutrition Plan to collaborate. But at the subnational level, sometimes they are not serious, and I have to push a bit. I always carry with me the NNP document with signatures.

PRIME respondents felt that there was good coordination at the national level, but they also stated a need for more intensive regional and subregional coordination. A woreda agriculture officer mentioned that there is a plan for multisectoral coordination for nutrition at the woreda level, but it is not being implemented. AGP-AMDe respondents mentioned that strengthening regional multisectoral working groups that include government and activity staff should be a priority, suggesting that ENGINE would be an appropriate entity to support the working groups, as it is already working well across sectors.

Coordination between and among Feed the Future Activities

ENGINE’s and AGP-AMDe’s central activity managers discussed a challenge related to the different designs, targeting, and approaches being used by Feed the Future activities across the zones of influence. ENGINE managers based in Addis Ababa noted that it is challenging for ENGINE to provide technical leadership and support for coordination on the ground that is appropriate to each activity. For example, Graduation with Resilience to Achieve Sustainable Development (GRAD) uses community animators rather than DAs; Agricultural Growth Program-Livestock Market Development (AGP-LMD) and AGP-AMDe provide inputs in the form of loans, while ENGINE provides them free to vulnerable households. Where there is no sectoral overlap, co-location is more feasible. For example, ENGINE covers AGP-AMDe woredas with health- and nutrition-specific interventions. The Mission- and ENGINE-led nutrition working group for Feed the Future implementers at the national level helps address these challenges.

The differences in approaches and targeting have led to some tension around the development of materials. AGP-AMDe managers mentioned that while they appreciated ENGINE’s initial DA training manual, they felt they needed to develop one that was adapted more closely to their target group of agricultural cooperatives. ENGINE managers would like to review all the DA training manuals being used by implementing partners to harmonize and combine them, but it is unclear how feasible or desirable that is for other stakeholders. PRIME activity staff at the national and regional levels said the ENGINE training manual is very good and has adapted it for pastoralist areas.

Respondents from all three activities mentioned that the joint field visits to implementation sites have been useful learning tools for all Feed the Future implementers because the mix of staff has introduced new insights. For example, through joint visits, managers learned that some of the beneficiary targeting for ENGINE and GRAD needed to be adjusted because some of the households receiving support were not the most vulnerable. Also, during a joint field visit, managers discovered that the zonal government counterparts responsible for gender needed nutrition training.

At the regional level, activity staff do not have much contact with other Feed the Future implementing partners. ENGINE and AGP-AMDe regional staff have agreed to hold quarterly meetings, but it has proven challenging for the two activities to prioritize the time for them. Again, this challenge could be addressed by activating the regional multisectoral working groups, with government and nongovernment actors participating.

Coordination between Development Agents and Health Extension Workers

All activities are training DAs to promote nutrition, both in nutrition-sensitive areas, such as diversifying production and using increased incomes to purchase a diverse diet, and in nutrition-specific areas, such as ensuring consumption of diverse diets and improving hygiene practices. Given the significance of this role, it seemed critical for the review to explore how DAs worked with other nutrition actors at the community level, particularly HEWs. Regarding coordination and complementarity between the roles of HEWs and DAs, responses—whether from government officers, project staff, or DAs—relayed notably consistent views. The following selection of quotes from ENGINE, PRIME, and AGP-AMDe staff, government counterparts, and DAs working with all three activities demonstrated this similarity in perspective.

DAs focus more on the production of fruits, vegetables, and small livestock and promoting dietary diversity. 
- ENGINE staff

When DAs are promoting nutrition, they don’t go into breastfeeding, complementary feeding, etc. 
- PRIME staff

DAs deal with fruit and vegetable production; HEWs focus on demand creation for them. 
- DA Bore woreda

HEWs also teach about hygiene and sanitation. We help each other. When we go to the kebeles, we are covering the HEWs’ activities; they also cover for us. 
- DA, Amhara region, interviewed in Bahir Dar

PRIME respondents felt that a key responsibility of DAs was to work with community members to improve productivity of cattle by growing fodder and using improved breeds. Respondents expressed the belief that increasing the availability of milk and meat would automatically result in improved consumption of them within households.

For all three activities, the area of overlap between DAs and HEWs centers around cooking demonstrations, which they coordinate and hold together, usually at FTCs or schools. Some respondents referred to having regular meetings, joint planning, and joint review sessions between the MOH and MOA to further integrate the work of HEWs and DAs, but this seemed more aspirational than concrete.

Coordination with Other Community-level Actors

In all focus group discussions and interviews, reviewers probed about whether there are other nutrition actors at the community level who should be engaged in improving maternal and child nutrition. Respondents generally could not identify other actors. When asked specifically about whether private sector actors can contribute to improved nutrition, respondents at ENGINE were skeptical, explaining that the profit motive of private sector actors may be incompatible with public goods such as nutrition—particularly for vulnerable households. Even though ENGINE staff members were cautious about engaging with the private sector, ENGINE does work with local private growers to source hybrid chicks and with private input dealers to source some hybrid seeds. Given the experience that PRIME has in engaging private sector actors for development, PRIME staff members were mostly positive about for-profit actors’ potential to contribute to improved nutrition. Staff members were enthusiastic about the potential to leverage private sector resources to generate demand for diverse diets and to facilitate access to improved technologies for post-harvest processing and storage. Activity staff and a woreda agriculture officer felt that more large-scale food growing companies could be drawn to Amibara woreda, where large-scale cotton growers are currently doing well. However, there were no specific ideas on how to make the role of private sector stakeholders better oriented to support nutritional outcomes, (i.e., to be more nutrition sensitive).

Other nutrition-oriented actors or programs mentioned by DAs included Bahir Dar University’s CASCAP program (acronym was not known) and the international corporation DSM, which is working with schools in Bore woreda to assist with training on nutrition in schools.

Process

This section discusses findings related to key process components in the start-up, training, supervision, monitoring, and reporting associated with supporting DAs to undertake nutrition-related actions.

Start-Up

AGP-AMDe started before ENGINE, but when the AGP-AMDe team was planning the activity, they knew ENGINE would have technical assistance and coordination roles for nutrition. However, AGP-AMDe, needing to begin work and challenged by delays and the absence of a model for developing messages and materials, adapted a training curriculum and counseling materials from the SMFM model and Save the Children’s Cost of Diet tool. An activity manager explained:

… there were no other models or experience. Nobody had trained Ministry of Agriculture staff in nutrition. They’re anxious to work but don’t know how to promote nutrition.

AGP-AMDe also encountered challenges in engaging the MOA at different levels in different regions in support of nutrition.

When we approached them, they were a little surprised. We needed to start with advocacy work. We had to tell them [sic] importance of integrating. The heads of the federal and (SNNPR) regional HEA program both came to the initial HEA training in Awassa and gave speeches. We had to convince each region. We submitted our module, and they reviewed and gave permission. The document is important to them, and we have scheduled together to roll it out. 
- AGP-AMDe project manager

For PRIME, nutrition was not in the original proposal, but it was included by the time the first work plan was developed. Mercy Corps, PRIME’s lead implementer, had experience with another project that also utilized DAs for grassroots development, which PRIME drew from to develop the work plan.

Training

ENGINE is the first activity in Ethiopia to use this approach of promoting nutrition through DAs. The activity first drew from learning from USAID’s Infant & Young Child Nutrition (IYCN) (2006–2012)⁵ project’s urban gardening model, adapting IYCN materials until ENGINE could develop its own. However, ENGINE managers realized that without training on nutrition, woreda-level health and agriculture managers would not know how to plan and manage nutrition activities or how to also support DAs. In response to this experience, ENGINE produced a nutrition-sensitive agriculture training manual for program managers. Only ENGINE is currently using this manual, but it has recently shared it with other Feed the Future activities, which will need to adapt it to their activities. The DAs receive a two-day training on the importance of producing and consuming a diverse diet, the importance of the first 1,000 days, and instructions on how to do cooking demonstrations, among other nutrition messages.

We did a rapid assessment on nutrition knowledge gaps with DAs. They knew nothing about nutrition ... We trained and helped to integrate nutrition issues into their daily activities. We developed a training manual but included a lot of nutrition-specific messages. We thought if men are one of the contact points for DAs, they can learn more about nutrition-specific practices to be able to support their wives better ...
- ENGINE manager

Two DAs who have been implementing ENGINE livelihood activities for a while, including a nutrition-sensitive agriculture activity, described the changes in their work since being trained and supported by ENGINE. They mentioned that while their roles and responsibilities remain the same, there are changes in how their day-to-day activities are carried out.

We have the same roles and responsibilities. We worked with model farmers; now we work with vulnerable households and work more with women but still with model farmers, too.
- DA, Bore woreda

A woreda-level agriculture officer agreed.

ENGINE came here to start the discussion. They raised awareness of the problems of stunting and micronutrient malnutrition. Integration is a must; there is no more running alone. All have agreed on this modality of using DAs.

Training of trainers session with AGP-AMDe. Photo credit: SPRING

For AGP-AMDe, BCC and nutrition staff from the United States supported the development of the messages, training curricula, and job aids. They conducted a cost of diet analysis, as well as focus group discussions and a barrier analysis for key behaviors within each of AGP-AMDe’s value chains. Questions considered in developing messages included: “Do maize farmers need different messages than coffee farmers? What are the differences between men and women, between households that grow or buy vegetables in terms of behaviors and messages?” AGP-AMDe managers felt that ENGINE’s DA manual was not tailored enough to the agriculture sector, so they created training and SBCC materials, ensuring consistency with ENGINE’s messages.

AGP-AMDe faced some challenges when Regional Agriculture Bureau staff wanted to include more detail around nutrition; staff wanted longer and more technical sessions. Activity staff tried to limit content to higher-level concepts related to male support of nutrition behaviors in the household. The idea is that the HEWs, who counsel on nutrition more in-depth at the household level, will reinforce the behaviors that the DAs are promoting and vice versa. For example, while there are messages about exclusive breastfeeding (EBF) in the SMFM manual, they were included to convince male farmers of the importance of EBF so that they in turn support their wives to practice EBF.

To develop training materials, PRIME both pulled from the RAIN (a preceding activity) postharvest manual and adapted the initial ENGINE nutrition-sensitive training manual for DAs to their pastoralist context. The DAs trained by PRIME trained community members in Amibara woreda, Afar. It is a practical, three-day training, focusing on grain preparation, processing, and storage. It is also linked to the underground grain storage technology being promoted by PRIME through the DAs. PRIME also held a training on drug handling and quality for female pastoralists to improve livestock productivity and health. Again, PRIME assumes that improved production of milk will lead to increased availability of milk and milk products, increased consumption by women and children, and ultimately, improved household nutrition. PRIME is also exploring a training of trainers with farmer groups so that the groups can assist in scaling up of technologies (e.g., grain storage bags) and messages.

Supervision, Monitoring, and Reporting

ENGINE managers based in Addis Ababa, zonal livelihood coordinators, and government counterparts from regional and woreda offices participate in post training follow-up and supportive supervision. There are two types of supervision: integrated supervision, when government counterparts from different sectors work together, and internal supervision, when only ENGINE staff participate in the supervision. The ENGINE team developed a supportive supervision checklist that is used during visits.⁶ Activity managers conduct quarterly spot checks with regional and zonal teams for building the capacity of field staff.

ENGINE’s routine activity monitoring and reporting feed into the MOA monitoring system. There are no nutrition indicators in the MOA formats, so some parallel reporting for management and reporting to USAID is necessary. Activity monitoring of DAs focuses mostly on outputs, such as which households have received which productive inputs and training, how many cooking demonstrations were held, and how many community members participated. Nutrition indicators have not been well integrated into government monitoring systems across sectors, and there does not appear to be a mechanism for sharing monitoring data with communities.

For AGP-AMDe, supportive supervision of DAs promoting nutrition has not occurred, as the activity is just rolling out. A regional nutritionist has recently joined the team in Amhara. The nutritionist will spot check and attend DA trainings, which will be conducted by HEAs, who were trained as trainers with the SMFM curricula. The activity has also trained subregional team leads, so they can help with supervision.

AGP-AMDe wants to ensure that the training and implementation of SMFM is “scalable, trackable, and accountable.” The activity will use hard copy formats to follow up with farmers who have been trained. Until supervisors verify that DAs have trained the men and women they are responsible for training (20 male cooperative members and their wives), the DAs do not receive payment nor do they obtain their full certification. AGP-AMDe is still developing a monitoring system to capture behavior change, possibly using Lot Quality Assurance Sampling. The activity is using the cost of diet analysis to monitor whether incomes are increasing enough to meet the affordable diet gap. For both nutrition and agriculture trainings, SMFM concentrates on three messages, each with associated indicators, based on activity experience that promoting more than three practices at a time is not conducive to adoption. For example, DAs promote dietary diversity for the family, and the activity is monitoring vitamin A-rich food consumption to track behavior change related to the dietary diversity message.

For each activity that DAs are engaged in with PRIME, activity mangers based in Addis Ababa give approval, and the Woreda Agriculture Office assigns supervision and technical expert support. Field days for DAs reinforce new nutrition skills and knowledge. As with the other two activities, monitoring and reporting data are passed from the DAs to the woreda office, where they feed into activity and government information systems. PRIME field staff expressed interest in collaborating with a university to conduct a randomized controlled trial to determine the impact of adopting the underground grain storage technology on household food security and nutritional outcomes.

Challenges

Grain storage. Photo credit: Mercy Corps

Respondents from all three activities—whether activity managers, field staff, government counterparts, or DAs—highlighted similar challenges related to their roles in supporting DAs’ inclusion of nutrition messages in their extension work.

Within the ENGINE and AGP-AMDe implementation areas in Amhara, there are significant economic and cultural barriers preventing households from acting on the recommendations and new skills being transferred by the DAs. Several of these barriers are discussed below.

• The practice of fasting (not consuming any animal source foods) is rigidly enforced by church leaders and community members. Families fear judgment from their neighbors and ritual pollution caused by having animal source foods in the house during fasting days or longer periods such as Lent. Thus, even though pregnant and lactating women and children younger than seven years are not officially required to fast, women often fast anyway due to social pressure. Women who decide for themselves and their children not to fast may not have access to meat, eggs, or dairy products at home. This has a larger market effect of greatly reducing demand for animal source foods during fasting periods, so producers cannot obtain a fair price. Another barrier to dietary diversity respondents noted is that most groups in Ethiopia base their diet on one staple food, and people believe that “if the stomach is full, they consider that someone has good nutrition.” (ENGINE manager)

• Discussions with community members in Amhara and Afar underscored the reality that nutrition practices take place in a complicated context with competing priorities. For example, one woman in Bore woreda producing fruit and vegetables with the support of ENGINE explained,

  We can’t avoid selling, as we need money. The decision is based on the relative cash value of the food; for example, if we have cabbage and carrots, we’ll sell the one that will bring more money. I would like to base my decision on the nutritional needs of my child. Perhaps only households that are better off can do that.

• Another challenge was found in Amibara woreda in Afar region, where pastoralists face food insecurity caused by a loss of grazing land for their cattle and significant water shortages for both consumption and irrigation. These natural resource issues prevent community members from acting on DAs’ recommendations of producing animal fodder, cultivating vegetables for sale and consumption, and/or improving post-harvest processing and storage, particularly for maize.

Implementation and enabling environment challenges also may affect the use of this strategy, including:

• Challenges mentioned by male and female community members participating in PRIME and ENGINE included unclear communications from activities about which community members would receive which training, inputs, and other support such as subsidized access to grain storage technology. Community members also expressed an interest in visual aids or reminders to help them remember specific agriculture or nutrition behaviors or information.
• While nutrition has gained a higher profile through the efforts of Feed the Future, other USAID investments, and the SUN movement,⁷ respondents noted that political and cultural gatekeepers do not prioritize nutrition enough to bring about large-scale change. This applies to government officials, from parliament and national ministries to woreda offices, as well as to religious and community leaders and decision makers at home, who are generally men. Nutrition activities are rarely incorporated into regional and woreda workplans or budgets—whether ministry budgets or household budgets. A female DA in Amhara stated simply, “Sometimes nutrition requires spending money; this needs attitude change.”
• Activity staff, government officials, and DAs all noted challenges related to the DA system itself. DAs are often overloaded with multiple tasks and messages. Large coverage areas, lack of transport, and poor infrastructure make movement of personnel and inputs difficult. Respondents mentioned that the incentive structure for DAs does not sufficiently support their travel or time. High turnover of DAs, through resignation or government-mandated rotation, undermine the activities’ objective of building capacity of DAs to promote nutrition.

Successes

Although the three Feed the Future activities are still in the early stages of using DAs as change agents for nutrition, a number of initial successes emerged during the review. The successes center around ENGINE being established as a credible partner in the eyes of the Ethiopian Government; training for the DAs that imparts new skills; design flexibility, which allows activities to adjust to field realities; and the informal sharing of messages and practices among communities.

During the course of the review, ENGINE convened a meeting with national parliamentarians about nutrition. Staff members across the activities were excited about the meeting and the growing commitment of the parliamentarians to make nutrition an urgent priority in the national policy agenda. Staff saw this as an important way to continue progress, with different sectors following up on their commitments under the National Nutrition Plan. ENGINE’s leadership and technical assistance were mentioned positively by staff and government counterparts across activities. One AGP-AMDe manager commented:

We work closely with ENGINE and consider them as important a partner as USAID.

Respondents felt that the trainings implemented to date with DAs and other MOA staff have been successful at imparting new skills as well as changing attitudes about nutrition and its importance.

Already, with government officers and DAs, the buy-in and excitement about how we’re linking agriculture and nutrition has been very fulfilling. We’re getting great feedback on the trainings, and we hope the passion will roll down.

Respondents across the activities said word about the nutrition-sensitive training for DAs and activity managers is spreading; they are receiving requests for training from zonal- and woreda-level government officials outside of their implementation areas. Further, the design and planning are fairly flexible, which allows the activities to adjust to the realities on the ground. For example, PRIME began with an incremental step: improving post-harvest storage techniques. This approach has been well received because many households transitioning out of pastoralism have only known of maize as an alternative food crop. This population is principally concerned with food availability and growing enough food to last the year. Starting with improving postharvest storage to save more of a limited maize crop therefore helped quickly engage the target beneficiary group.

Community members participating in ENGINE and PRIME spoke very positively about what they have learned. Women related to producing and consuming different kinds of food, especially adding fruits and vegetables to their diets, which was bolstered by recipes demonstrated at FTCs or schools.

Instead of giving single food, we have started introducing vegetables and some other things based on the cooking demonstrations we saw. We are trying to add to our children’s foodstuffs. 
- Woman, Bore woreda

The male respondents focused a bit more on the success of fruit, vegetable, and small livestock production in ENGINE implementation areas. They also focused on the success of the maize post-harvest processing and storage activity in the PRIME implementation area.

Finally, community members are sharing what they learn with their neighbors, both informally and at cultural and religious gatherings. In Bore and South Achefer, the fact that DAs are promoting women’s participation is important. A woman in Bore told reviewers that before ENGINE started, community members did not hear anything about nutrition, but now they meet regularly with DAs and HEWs and are sharing the information with other women in the community. There is also intergenerational sharing happening, pointing to the success of the strategy of DAs engaging teachers and schools in gardening and cooking demonstrations. When asked whether families are practicing what they learn at cooking demonstrations, one father in Bore said proudly, “The teachers give some training about nutrition to students. The children come home, and they demonstrate for us. The students are the mastermind for that; they are the equivalent of DAs when it comes to ingredients and how to prepare food.” This bodes well for families taking nutrition more seriously when it is promoted through formal education and for the future families of these children.

Lessons Learned/Advice from Respondents

Although the time for integrating nutrition in the activities of agriculture DAs has been short, a number of lessons have surfaced from this process review. Several respondents have offered advice for activity designers and implementers.

Quality Trainings: DAs and government counterparts who have been trained by all the activities felt the trainings were participatory and useful to them in their positions. Several respondents suggested that refresher trainings and repeated demonstrations are important for implementers and community members to retain new knowledge and skills: “No one can learn anything at one time” (DA, Amhara region). Some respondents felt that more in-depth training around nutrition would be helpful, while others felt that would be counterproductive, taking away from DAs’ ability to effectively deliver a core set of nutrition messages and skills. Activity managers stressed that experience to date has taught them to focus trainings as much as possible. They also stressed that it is important to give the DAs tools to train others by teaching them how to plan and giving them reminder materials or teaching aids.

Appropriate Social and Behavior Change Communications: Field personnel stressed the importance of appropriate SBCC. Cultural traditions must be adequately assessed. Posters and illustrations assist in understanding, but more importantly, communities adopt the prescribed behavior when they see tangible results. Feed the Future implementers should work on common approaches to SBCC beyond common messages. Activities need to assess the context, especially the cultural traditions of target populations, and develop appropriate and creative SBCC materials that incorporate visuals as well as text. Designers and implementers need to understand stakeholders and their motivations, as they may be different for farmers and mothers, for example.

Consistency and Follow-Up: Community members stressed the importance of consistent and sustained engagement and communication among activities, government, and communities. A woman in Bolta said, “The only time we see people is when people like you come writing on papers. If you are willing to help us, we are ready to learn,” while a male respondent said, “Please always come and check on us.”

Multisectoral Coordination: Most people agreed that nutrition requires a multisectoral approach and that coordination, especially at the sub-national levels, needs to be strengthened. “People think multisectoral coordination is only about meetings. It is about understanding different sectors’ comparative advantages. Look at how each activity is using DAs and be realistic about what they can implement well” (ENGINE manager). The different types of community-level workers may be leveraged, necessitating joint training and planning at the woreda level.

Monitoring Behavior Change: Activity managers recognized the importance of strong monitoring and reporting that capture changes in behavior as well as service and input provision and knowledge acquisition. All three activities are still learning how best to do monitoring, which is essential to robust quality assurance via supportive supervision.

Scaling Up and Down: Interventions such as cooking demonstrations, mentor households, and model farmers intend to scale up nutrition messages to a greater number of households. They also bring the “cascade” closer to communities while expanding to other communities. Here, effectively reaching of community influencers (religious leaders, household heads, policymakers, clan leaders, and agricultural research centers) is key, as are the resources and support for proper research and documentation.

Gender

Gender emerged from the review discussions as a cross-cutting theme, touching on various aspects of implementation and coordination. At first, staff from all three activities, government counterparts, and DAs asserted that gender was not a big issue as far as DAs promoting nutrition went. DAs are mostly men dealing with mostly male clients, and HEWs are mostly women dealing with mostly female clients. But when the issue was discussed further, it became clear that the gender of DAs does matter, and that the way they work and the way they are perceived by clients is influenced by well-established gender roles in society. As one agriculture officer in Bore noted,

There is no problem in the case of gender affecting DAs’ roles. It is known that women are more active in cooking, so for cooking demonstrations the women participate. Some male DAs can cook, but aren’t talented in that area.

A staff member in Amhara went so far as to say:

Female DAs for women is a better idea ... and [women] raise their internal issues with female DAs.

A community member in Amibara, Afar, where PRIME works, also commented:

It would be easier and more comfortable for women DAs to enter the households and teach regarding nutrition.

Carrying goods in Amhara Region. Photo credit: SPRING

There is consensus that while women generally manage the direct food preparation, feeding, and caring, men’s influence is very important for decisions about what to produce, what to consume versus what to sell, what foods to purchase, and whether and when to spend money on health care. Respondents were optimistic about the ability to influence men around these issues if DAs discussed and demonstrated improved practices with them.

Respondents acknowledged challenges with increasing the participation of women in co-ops and other agriculture groups. AGP-AMDe has responded to this challenge by asking DAs to train 20 farmer members along with their wives while also increasing the number of women co-op members.

In addition to increasing women’s participation and access to services, respondents stated that it is challenging to improve women’s control of resources. The types of resources that women have some control over are poultry, homestead garden, and petty cash work. But there are even challenges with improving their return from these resources. An ENGINE manager noted that as soon as the small livestock intervention began to generate income, men wanted to take control of the activity.

Respondents did not independently mention the issue of women’s time or energy expenditure, but when probed around these issues, they mentioned that the activity found this to be a problem for the gardening, but not for small livestock. The activity revised the vegetable package it was promoting (removing lettuce) to reduce the need for water to address women’s time and labor. Reviewers shared with activity staff that women in the community focus group discussions said that it is common for men to make their wives work before the traditional 40 day rest period after delivering children has passed and requested project support to cope with this. In response to this information, an activity manager commented that ENGINE’s SBCC strategy is promoting extending the period of post-delivery rest to at least the full 40 days.

When respondents were probed about the evidence of the very low status of women at the activity sites visited, they were open about the challenges the situation presents. For example, married women in the target population are expected to keep their hair shorn and go barefoot out of deference to their husbands. In addition to the strong tradition of male dominance, there is a custom of placing a high value on maternal selflessness. An activity manager explained,

Women never eat alone; they won’t cook for themselves. Others will say she is selfish. Selflessness is a value.

Improving the status of women and improving the care and support mechanisms for women are two reasons ENGINE is engaging the assistance of the Orthodox Church. Another reason is the very strict enforcement of fasting rules, which do not apply to pregnant and lactating women or children under seven years old. The Church is important in bringing about social change around gender roles and issues of fasting. For example, families believe that one cannot cook meat at home during fasting because it will contaminate the cooking utensils, even for the household members who do not eat the meat. People do not cook meat even if there are small children, women, or ill people at home because the neighbors may notice the cooking smells and say they are not fasting.

After probing about whether men in the community accept the nutrition-related changes that the DAs are promoting, interviewers received mixed answers. One woman said,

It depends on the mindset. Some men like to see their wife and children being well off.

Others disagreed. The women said it varies from household to household. When asked whether more positive men can influence the more negative ones, another woman responded,

If a couple is fighting, the elders may come to arbitrate. Everyone will say okay and kneel down while the elders are there, but when they leave it may continue as before. Some do change practices, though, and there is more room to change.

However, the women agreed that there have been positive changes for women since the activity started. One respondent said,

Thanks to this activity, we are more empowered. We appear in public with our male counterparts. The DAs and HEWs call us to meetings and we participate. We are in a better position than before.

This indicates that the activities’ use of women to engage women can help address challenges of unequal access to agriculture extension services and technologies.⁸

Conclusion and Recommendations

Consultation in Afar Region. Photo credit: SPRING

This review examined the way in which three Feed the Future activities in Ethiopia use agriculture extension workers to promote maternal and child nutrition in agriculture programming. Areas explored were built upon the learning from SPRING’s earlier global review of Feed the Future programming titled “Leveraging Agriculture for Nutritional Impact through the Feed the Future Initiative: A Landscape Analysis of Activities Across 19 Focus Countries.” This review suggested an opportunity to investigate the coordination within and between Feed the Future activities around using extension workers to deliver nutrition messages and interventions. It also enabled documentation of the processes being used in this programming approach, as well as an initial assessment of challenges and lessons learned at an early stage in implementation.

This review was not an evaluation of the programming approach or its merit relative to other methods of achieving behavior change for nutrition outcomes. Nevertheless, there are lessons that can be drawn from the work to date of the partners in Ethiopia. From these lessons, SPRING proposes several recommendations to improve nutritional outcomes through agriculture programming. Not every challenge or lesson learned noted in this report has led to a recommendation; rather this report presents the following ideas for consideration by activity designers and managers to enhance current agriculture interventions and to design more nutrition-sensitive agriculture investments in the future.

Context should inform initial activity design as well as mid-course corrections. The nutrition-related practices and skills that DAs are promoting will only have impact if they are feasible for people to apply in their daily lives. The review found significant economic and cultural barriers preventing cooperative members or household members from being able to act on or use the recommendations of the DAs. For example, in Amhara region, women have very low status. They do not participate in the groups targeted by DAs and have relatively little ability or voice to press for household investments in health and nutrition. Their workload is already among the highest in the country, yet DAs are recommending that they spend more time on cultivation, food preparation, and feeding and child care. In Afar, food is not available year round for poor and vulnerable pastoralist families due to serious lack of water and loss of cattle (a traditionally reliable food source). Recommendations by DAs to diversify production and consumption may not be effective or appropriate when target populations are primarily concerned with the challenges surrounding their transition out of pure pastoralism and diversification to sedentary farming. They are also primarily concerned with the challenge of meeting their needs for basic food and water. It is recommended that activities be encouraged to undertake nutrition-sensitive interventions that seek to address the basic causes of malnutrition, such as depleted environmental resources or harmful socio-ideological frameworks. To do this, ENGINE could be supported to convene a meeting of Feed the Future partners that are using DAs to promote nutrition to discuss specific messages and practices being promoted. Implementing partners could focus on their experience so far with contextual and cultural barriers that groups or individuals face when trying to act on the recommendations that activities are promoting. ENGINE could take lessons learned from this meeting to develop guidance for activities to tailor current messaging so it is more appropriate and effective. ENGINE could also use the learning to develop new messages focusing on contextual factors that the activity could help to change, such as gender norms or community planning for better access to water.

Challenges within the DA system can be addressed by building on recent political attention and prioritization being given to nutrition in Ethiopia. With nine ministries signing on to the National Nutrition Plan and parliamentarians committing to prioritizing nutrition within the national policy agenda, Feed the Future partners may be able to leverage greater attention from the MOA to the role that DAs can play in supporting nutritional outcomes. In Ethiopia, ownership of and accountability for nutrition policies and strategies are more likely to be sustained over time if they come from the top, and if nutrition-related activities are incorporated into government line ministry work plans and budgets. Widely recognized systemic challenges within the DA system include incentive structures, geographic coverage, work burden, gender representation, and turnover. These challenges can only be addressed in partnership with government. In its role to coordinate and support nutrition interventions across the Feed the Future portfolio, ENGINE is particularly well placed to advocate on behalf of implementing partners and can highlight their nutrition activity-related efforts and needs with key ministries from the national to the woreda level.

As embodied in the National Nutrition Plan, mechanisms to support cross-sectoral collaboration and increase leveraging of resources should be strengthened at the national, regional, zonal, and woreda levels. It is widely recognized that improving nutritional outcomes requires a multi-sectoral approach. Grassroots workers integrate activities where they can, but joint planning, monitoring, and learning among implementing partners at the subnational levels and including public and private services wherever possible could help make both food and health systems more effective. Stakeholders felt that ENGINE has been playing an important role in supporting cross-sectoral collaboration and suggested that it could play an even larger role, particularly at the regional and zonal levels. While ENGINE’s design incorporates nutrition in the health and agriculture sectors, stronger linkages with WASH, health, and education are also needed.

Implementing partners can improve the nutrition-sensitivity of their activities by applying the agriculture-to-nutrition conceptual pathways during annual work-planning and/or design of mid-course corrections. Again, ENGINE is well placed to provide technical support for this process. Examples of the types of actions that may result from leveraging connections across and along the pathways include:

• Increasing efforts to reduce women’s labor and time burden by improving access to technologies while at the same time potentially reducing post-harvest losses.
• Identifying and strategizing how to fill gaps if interventions leave out steps along the pathways. Steps that are commonly overlooked in both activity design and implementation approaches include women’s energy expenditure (labor) and time burden, equitable household decision-making and allocation of resources for health and nutrition and monitoring the effect of local market prices on purchase and consumption behaviors, especially among vulnerable households.
• Supporting selection of appropriate intermediate indicators to monitor progress and gauge whether interventions are on track to meet planned nutrition outcomes.

ENGINE has an important role to play in continuing to work with the Mission on developing comprehensive SBC guidance. This guidance is under way and is attempting to incorporate findings from implementing partners’ formative research and monitoring data, and to draw on the evidence base related to Essential Nutrition Actions. SPRING recommends that partners use the guidance as a “menu” of proven approaches regardless of differences in objectives, targeted groups, and implementation contexts. ENGINE can then help partners develop effective SBC approaches, priority nutrition messages, and communication materials. This should continue to be a priority across the different Feed the Future activities as shared SBC guidance and strengthened capacity to design and deliver appropriate nutrition messages will likely enhance nutrition outcomes.

Not all agricultural interventions should target nutrition outcomes at the household level. Differences across activity design, objectives, and approaches result in different comparative advantages for the activities. In fact, some activities, especially those focused on larger-scale value chains, may be better placed to create change at a systemic (e.g., food systems) level. Separation of roles within the Feed the Future portfolio, with strong coordination among regions and/or woredas that comprise the zone of influence, could ensure that Feed the Future activities are operating in a more complementary way. In this way, the Mission would simultaneously promote change at the household, community, and food system levels, focusing each activity on the effective interventions that have the most potential.

In closing, this review enabled the documentation of a process that is still very much in progress. It is meant to provide current insights, from the perspectives of different stakeholders. This may help to adjust current programming and coordination and generate ideas for future activity designs and harmonized intervention approaches. It does not provide a final word on integrating nutrition into the work of agriculture extension agents; but it is meant to inform ongoing conversations among Feed the Future practitioners and implementers about the relative merit of this programming approach, and what constitutes better practices for designing, implementing, and monitoring this approach.