An analysis of Fresh Food Voucher Programme piloted in Ethiopia

By Pankaj Kumar, Anne Marie Mayer and Elizabeth Molloy

Pankaj Kumar has been working in Ethiopia with Concern Worldwide since 2010. Previously he has worked in Zambia, Zimbabwe, Bangladesh, Nepal, Liberia and other African countries. He has keen interest in agriculture and nutrition issues.

Anne-Marie Mayer has been working with Concern Worldwide for the past 2½ years as a Food and Nutrition Advisor supporting programmes in several African countries. She has a PhD in International Nutrition from Cornell University and has worked for NGOs and academia on the links between agriculture and nutrition.

Elizabeth Molloy has been working with Concern Worldwide in Ethiopia since November 2012. She is an education and gender specialist who has previously worked in The Gambia and Nepal. She will be moving to Malawi in May 2013 to take up the position of Concern Worldwide’s Education Programme Coordinator.

The authors would also like to thank Suzanne Fuhram, Sosena Mellese, Nardos Birru, Mesfin Kibret, Terefa Getachew, Charlotte Walford, Martha Yigezu and other staff from Concern’s Sodo office for their assistance in providing information and reviewing this article. Additional thanks also go to all Kindo Koysha Woreda officials, Health Extension Workers, Outreach Workers and Voluntary Community Health Workers for their dedicated work to make this project a success. We also thank the beneficiaries and non-beneficiaries for sharing their insights with us.

This article describes Concern Worldwide’s experience implementing a Fresh Food Vouchers programme (FFVP) linked to a targeted SFP programme (TSFP) in Wolaiyta Zone of the Southern Nations Nationalities and People’s Region (SNNPR) of Ethiopia in 2012. The FFVP was implemented within a larger programme aimed at promoting longer term food and nutrition security.

Background

Wolaiyta zone¹ in SNNPR of Ethiopia is one of the most densely populated zones in the region with a population of 1,792,682. It consists of three agroecological zones (arable highlands, midlands, and lowlands) and is one of the most vulnerable and famine-prone parts of the region due to its chronic poverty and food insecurity caused by a combination of factors. These include high (and increasing) population density, diminishing landholding sizes, intensive cultivation leading to soil infertility, periodic rain failure, crop pests, livestock diseases, and lack of alternative income sources.

In 2011, 86,359 (53,651 male-headed and 32,708 female-headed) rural households in Wolaiyta zone faced critical food shortages for more than six months, and dependence on the government’s Productive Safety Net Programme (PSNP) - a largely cash based form of social protection - was very high. The major coping mechanisms of these vulnerable communities included PSNP, begging, eating unpalatable wild fruits, and daily labour. Rural livelihoods, especially of the extreme poor, are often vulnerable to risks and shocks.

By the end of April 2012, SNNPR had received only 25 to 50% of its expected rainfall which led to a very poor harvest and resulting food shortages. Kindo Koysha is one of 13 woredas in Wolaiyta zone and is extremely vulnerable to malnutrition. It was ranked a ‘no.1 hotspot’ woreda by UNOCHA in 2012. A Concern WW baseline survey conducted in December 2012 found that the average household ‘food gap’ in Kindo Koysha was seven months. It also reported that 14% of households with children below 5 years of age in the woreda had children admitted to the TSFP in the preceding year, while 15% of pregnant and lactating women (PLW) were admitted to TSFP in the same period. Previously part of Damot Weyde, the woreda has a history of significant malnutrition prevalence; in the last ten years, Concern WW has responded through ‘emergency nutrition’ interventions a number of times.

Concern WW’s operations in Kindo Koysha woreda

In July 2012, Concern WW began supporting Community-based Management of Acute Malnutrition (CMAM) in five woredas: Kindo Koysha, Kindo Didaye, Damot Gale, Dugna Fango, and Damot Weyde in SNNPR as part of an emergency response. This was prompted by deterioration in food security, and an increase in malnutrition prevalence rates (detected through screening on Child Health Days and an increase in the number of OTP children). Consequently, the programme targeted children below five years of age and PLW. In Kindo Koysha, this included setting up a TSFP at 20 sites, supporting government outpatient therapeutic programme (OTP) services at 26 health centres and health posts and stabilisation centre services (SC) at four hospitals, and supporting community screening and referral of acutely malnourished children and women. Concern WW focused on capacity building of woreda health staff to implement national CMAM protocols, as well as facilitating joint supervision of CMAM sites and providing logistical support.

Kindo Koysha was selected for the pilot of Concern’s FFVP. In addition, an Infant and Young Child Feeding (IYCF) intervention was implemented in Kindo Koysha (and Dugna Fango). This ‘Smart and Strong Families’ IYCF initiative aims to empower mothers to become ‘model mothers’ capable of promoting positive IYCF and family nutrition messages throughout their communities. Cooking demonstrations and health education sessions are used to promote positive IYCF practices. These sessions were held at village level reinforcing IYCF messages at household level.

Most recently, through ECHO funding, Concern WW has initiated a resilience project in Kindo Koysha, which comprises a multi-sectoral approach to poverty reduction involving emergency nutrition, livelihoods, water, sanitation and hygiene (WASH) and disaster risk reduction (DRR).

The FFVP and related IYCF activities implemented in Kindo Koysha are described in this article.

Overview of Fresh Food Voucher Project (FFVP)

Concern WW’s FFVP in Kindo Koysha was initiated to supplement the foods provided by the TSFP and to compensate, to some degree, for the sharing of the TSFP cereal ration at household level, while improving caregivers’ knowledge and skills in using local foods to promote child nutrition over the long term. While the TSFP ration is calculated to provide a nutritious supplement to a child’s family diet while he or she is recovering from moderate acute malnutrition (MAM), it was apparent that sharing of Corn Soy Blend (CSB)/Famix and oil among family members was widespread and this was affecting recovery rates. Complementary feeding practices were also observed to be fairly poor in the area, due to poor access to nutritious foods, as well as insufficient knowledge and skills on how to prepare them. The FFVP offered a practical means of exposing mothers to nutritious local fresh foods while actively demonstrating how to incorporate them into their children’s meals.

The aim of the project was to contribute to a reduction in mortality, morbidity and suffering associated with MAM amongst children aged 6- 59 months and PLW in the target area. This was to be achieved by improving dietary diversity of target beneficiaries through the provision of FFVs, which were exchanged for fresh fruit, vegetables and eggs at weekly distributions. The project was scheduled to run for three months from August - November 2012.

A voucher scheme was considered more appropriate than a cash distribution to avoid conflicting with the established PSNP. The approach used enabled the programme to address the limited availability of fresh foods in local markets and poor existing knowledge of the importance of fresh fruit and vegetables and eggs in a child’s diet. Positive examples of similar schemes informed the approach. For example, Action Contre la Faim (ACF) has implemented FFVPs in differing contexts and has developed good practice guidelines based on these experiences.

The target groups for the project were children aged 6-59 months and PLW who were admitted to TSFPs in the target woreda. Based on the national guidelines, children were admitted to TSFP with mid-upper arm circumference (MUAC) between 110 mm and < 120 mm and PLW with MUAC < 210 mm. All children discharged from OTP were also admitted into the TSFP and thus the FFVP.

Process

The first FFVP distribution was made at the end of August 2012 at the same time as IYCF activities commenced. The beneficiaries received the standard TSFP ration², as well as a FFV. The CMAM 2012 intervention had begun 1 month prior to the FFVP as a new programme.

Prior to initiation of the FFVP activities, a market assessment (based on the Emergency Market Mapping and Analysis tool) was undertaken to assess available fresh foods in the community and to identify vendors who were already in the market and trucking foods to neighbouring towns for sale. Women’s groups that sold eggs in the woreda were also identified. Potential vegetables vendors could not be identified in Kindo Koysha itself, however vendors who came from Sodo (zonal capital - 1 hour away) to the main market in Kindo Koysha were able to supply vegetables.

Subsequent discussions with woreda officials and staff at the woreda Trade & Industry office and the Marketing & Cooperative office led to the formation of a project cooperative. This allowed credit backing of the vendors and registration with a Tax Identification Number (TIN). The downside was that the members had not previously worked together and the only way to engage with vending to the project was through the co-operative, limiting possibilities for smaller venders/suppliers.

A contract was drawn up with the vendor’s cooperative. In order to avoid overwhelming the existing market system, the cooperatives were contracted to provide the fresh foods at kebele (village) level. A schedule was created to ensure all kebeles were served on a weekly basis so that beneficiaries were able to access fresh food at regular basis, and efforts were made to match the weekly fresh food distribution with the fortnightly TSFP distribution. Most of the TSFPs were at the kebele level.

Vendors brought fresh foods to TSFP distribution sites on trucks or by other means, such as by donkey, when road conditions were poor. Fresh foods included eggs, mango, papaya, avocado, banana, carrot, beetroot, tomato, garlic, onion, pepper, dark green leafy vegetable, pumpkin, and oranges. Beneficiaries were provided with weekly coupons redeemable for 2 eggs, 4 kg of fruits, and 3.5 kg of vegetables. Ration size was based on the recommended daily intake for children aged 6-59 months and based on 2 to 3 children below 5 years of age per household, but also considered that sharing would occur within the household due to the prevalent food insecurity situation in the area. The ration for PLW was the same as that for children under 5 years. Table 1 demonstrates the ration size.

Table 1: Ration size at planning stage
	Type	Servings per day per beneficiary	Servings per month
Fruits	Mango	0.66 pcs	19 pcs
	Avocado	0.66 pcs	19 pcs
	Banana	0.66 pcs	19 pcs
Vegetables	Dark green leafy veg	200 g	6 kg
	Carrot	115 g	3.45 kg
	Tomato	155 g	3.45 kg
Animal Products	Egg	0.28 pcs	8 pcs

 

At distribution, beneficiaries were registered and received a voucher redeemable for a one week ration. The voucher was in Amharic and is shown in Figure 1.

During the fresh foods distribution, TSFP also took place that included health education sessions. These focused on basic nutrition, IYCF, good hygiene, care-seeking for illness, and other topics. Cooking demonstrations to explain how to prepare different recipes with the fresh foods also took place at each distribution site. Recipes showed how to mix and cook vegetables and egg with the CSB and oil rations. Additional messages were developed to address specific issues regarding fresh foods, such as how to store the foods correctly and how to wash and prepare the items. Some mothers believed that the items could not be consumed by children as early as six months of age, so messages were developed to address this specifically.

Vendors were responsible for weighing the items and for distribution. Quality and quantity issues were overseen by a Concern WW Outreach Worker. Some issues with the quality of products were witnessed at the beginning of the project, such as overripe fruit, or bruised vegetables. These issues were addressed and beneficiaries were also vocal in ensuring quality foods were received.

It was originally planned that beneficiaries would choose which fruits should make up their 3.5 kg and which vegetables should make up their 4 kg ration. However, this proved difficult logistically. During distribution, instead of beneficiaries choosing fresh food, vendors reported that in many cases, beneficiaries would ask the vendors to choose for them the most suitable foods, as they were unfamiliar with the foods available.

Cost of the fresh foods for the programme was on average 381 Ethiopian birr (17 euro) per individual per month. The cost included transport cost and other related costs, but excluded costs for Concern WW staff to monitor the programme.

Programme monitoring and evaluation

The programme was evaluated using qualitative assessment tools addressed to beneficiaries, non-beneficiaries, vendors, health extension workers, outreach workers, cooking demonstrator, officials at the woreda office of Cooperatives and Marketing, Concern WW staff and managers. Interviews took place during the final week of the programme (26 November 2012) in Wolaiyta.

TSFP performance indicators for children aged 6-59 months were within national (Ethiopian MAM guidelines September 2012) and international (Sphere Project 2010) standards for proportion of exits that were cured, died or defaulted (Table 2). There is no agreed national or international standard for average weight gain or average length of stay in a TSFP, but the figures are also provided in Table 2.

Table 2: Performance indicators for the TSFP for children 6-59 months
	Cure rate	Death Rate	Defaulter Rate	Non - responder	Average Weight Gain	Average Length of stay
Kindo Koysha	86%	<1%	<1%	13%	5.5g/kg/day	94 days

 

Monitoring of the FFVP involved baseline and endline questionnaires administered by Concern WW Outreach Workers to the caregivers of more than 250 children during their admission to TSFP and FFVP and following their discharge. The sample was evenly distributed across all five CMAM kebeles. The objective was to measure whether there was improved dietary intake (as measured by dietary diversity score) amongst the children who received support.

Individual dietary diversity score

Table 3 and Figure 1 summarise the findings on the proportion of children 6 - 59 months of age who consumed different numbers of food groups based on the baseline (at admission) and endline (at discharge) questionnaires³.

Table 3: Admission and discharge results for consumption different food groups
Food groups	Baseline (at admission) (n=307)	End line (at discharge) (n=260)
Up to 1 food group	29%	0.4%
1-2 food groups	48%	10.5%
2-3 food groups	53%	17.8%
More than 4 food groups	4.2%	71.4%

 

The mean dietary diversity score changed from 1.96 at admission to 4.17 at discharge and the proportion of children who received more than four food groups increased from 4.2% to 71.4%. Children who predominantly ate from one food group at admission, increased to eating three or more food groups by discharge, with increased consumption of dark green vegetables, egg and fruits especially. Consumption of eggs typically drops from July to November because eggs are usually scarce in the market and production decreases due to the onset of the rainy and cold season. Furthermore, the production of vegetables from homestead gardens usually drops after the rainy season as water is scarce and there are no irrigation facilities. Therefore the increase in consumption of these foods is likely to be due to the FFVP. Further, consumption of legumes and fruits are generally not part of the routine diet though they are available throughout the year in the area. Thus this increase in legumes and fruit consumption could also be attributed to increased availability and promotion of these products through the FFVP.

The observed increase in consumption of dairy products (that were not supplied in the FFVP) could be due to IYCF activities leading to increased awareness; an observation highlighted through different focus group discussions, field visits and key informant interviews.

Pulse consumption also increased and may be explained by the local harvest of haricot bean (a major crop) during the months of October and November. Changes in consumption of grain and cereals like maize and barley were not observed; they remained the staple foods of the area.

Lessons learned

Planning and M&E

There appeared to be some confusion among Concern WW staff and woreda officials regarding the specific purpose of the project. Various objectives were cited, such as to reduce incidence of malnutrition, to increase dietary diversity and to reduce micronutrient deficiencies. However, provision had not been made to measure whether all of these objectives were met. The project objectives therefore needed clarification and communication to all stakeholders, as well as monitoring systems put in place to measure their achievement.

Lack of good baseline data was problematic. For example, it was difficult to attribute changes brought about by the project, as the baseline and endline studies in Kindo Koysha took place in different seasons and this influenced food availability that was not related to the FFVP.

A comparison of TSFP performance data across woredas was initially considered a plausible method to evaluate improvements in recovery rates attributable to the FFVP. However this was deemed of limited value because confounding factors, such as coverage, start of programme and factors related to agroecological zones could not be controlled adequately in the comparison.

Administration

All interviewed stakeholders stated that the project had run smoothly. However, some difficulties were reported with recording procedures. The paperwork involved in distributions was time-consuming with some elements unnecessary. Complicated procedures increased staff costs and reduced the time available to outreach workers for follow-up visits and supporting beneficiaries. Staff suggested that a data manager be employed to help with the administrative tasks.

Vendor system

The two vendors interviewed were both positive about the project. Both vendors and officials interviewed stated that the project had led to increased production of fresh foods in the woreda.

Vendors and Concern WW staff reported that initially payment to vendors had been delayed by a number of weeks. Concern WW staff attributed this to the extensive verification procedure involved. Fortunately, access to co-operative savings meant that vendors’ credit rating was not damaged by the delay in payment.

Understanding the ways in which co-operatives function proved difficult. From discussions with woreda officials, it appears that there are co-operatives of producers, who then sell their produce to co-operatives of vendors. Whether poor farmers have an opportunity to become members of co-operatives and contr-i bute to the supply of fresh foods is not clear.

IYCF and cooking demonstrations

Beneficiaries reported finding IYCF education sessions and cooking demonstrations very useful in focus group discussions and were able to recall and describe correct breastfeeding and complementary feeding practices from IYCF sessions. Focus group participants also reported replicating recipes learned at cooking demonstrations at home. The cooking demonstrator also reported receiving positive feedback from beneficiaries that they were implementing what they had learned in their own homes.

Perceptions of beneficiaries

Prior to the introduction of the FFVP, few of the focus group participants regularly fed their families fresh foods. Both access and affordability were cited as barriers to eating fresh foods. Where fresh food was available (such as in Bele town, the woreda administrative centre) the cost was prohibitive, while in more rural, isolated areas, even if women could afford fresh foods, they were not available. Money is prioritised for staples when available. Beneficiaries do not produce fresh foods (although some have an avocado and/or mango tree) due to infertile/ lack of land and lack of other resources. In an earlier mid-term review by Concern WW staff, beneficiaries and kebele leaders expressed a lack of awareness in the communities about the importance of fresh foods for nutrition, stating that fresh foods were only for urban people and not for rural people.

All beneficiaries from all five focus groups (sample total =25) agreed that they had learned a lot from the programme, e.g. awareness of the importance of fresh food to health. All focus group participants reported that the FFVP had a beneficial impact on their own health (in the case of PLW) and on that of their child. The participants were positive about the quality, variety and amount of fresh food provided by the programme. Four out of five focus groups reported that they had no problem with the extra time spent on the FFVP, between attending distributions and the additional time to cook separate meals for targeted beneficiaries. One focus group indicated that they were very busy cooking, fetching drinking water and firewood and that the return journey to the distribution took 4 - 5 hours. Beneficiaries’ willingness to take the time to attend distributions and education sessions without complaint indicates the high value they placed on the project.

While all of the focus group participants expressed a desire to continue consuming fresh food and providing it for their families, some had reservations about their ability to do so. Four of the five focus groups cited expense as a barrier to access, while two of the focus groups in remote locations stated that fresh food wasn’t available in their locality. Poor growing conditions also hindered fresh food consumption, although two participants planned to begin growing their own fresh food.

Additional observations

The qualitative assessment received no reports of the project having negative unforeseen consequences on food prices in the market. None of the non-beneficiaries interviewed complained about their lack of inclusion or reported any adverse effects of the project on themselves or others.

The introduction of the programme was timely - only 1 month after the start of the TSFP. While the end point had been planned and communicated to stakeholders, all felt it should have been extended. This was a view expressed by all groups interviewed: beneficiaries, nonbeneficiaries, vendors, programme staff and officials.

All interviewed stakeholders reported being satisfied with the targeting of the project. Non-beneficiaries reported that they feel happy to see neighbours benefit and to see improvements in their children.

"We feel glad that we are excluded because this showed our children were OK and healthy".

Health Extension Workers and outreach workers noticed that the FFVP encouraged mothers to attend the TSFP. The FFV programme has altered perceptions about how malnutrition can be addressed in communities.

Lasting benefits reported by beneficiaries included learning on hygiene practices, IYCF and from cooking demonstrations, with some starting to teach the topics to other community members. Some of the foods provided allowed beneficiaries to save seeds to plant on their own land, e.g. pumpkin seeds.

Vendors had started to view local and distant markets as potential areas for future selling of fresh foods but were concerned that beneficiaries would be unable to afford to buy fresh foods after the end of the FFV programme.

Conclusions and recommendations

The FFVP was well received by communities and the beneficiaries reported health and nutrition benefits for their children above those from the TSFP alone. The awareness of importance of fresh foods seems to have been raised by the project - not just by the beneficiaries, but also the wider community, vendors, health workers, Concern WW staff and local government employees. This is an additional benefit and more lasting aspect of the project.

During an emergency, the FFVP is a good model for introducing fresh foods to communities that do not have money or access to markets. The project works well alongside a TSFP, which provides for energy and protein requirements of the diet. Such a project, which sources local foods for distribution as part of surge response, also has potential to provide longer lasting social and economic benefits than a distribution sourced externally.

The CMAM data was difficult to interpret as part of the evaluation because there were potential confounding factors present in the different woredas that could not be assessed adequately to control for their influence on the results. The information on cure rate, default rate, average length of stay and weight gain could be misleading. In future interventions, greater attention should be given to an evaluation design at the planning phase.

Recommendations emerging from this experience include:

• The Behaviour Change Communication (BCC) component should continue after the end of the surge response and include all groups in the community. The improved practices would build community resilience so that during an emergency, communities would have at least the awareness and understanding of good IYCF and hygiene practices. A barrier analysis should be carried out so key determinants of behaviour are understood before implementing BCC activities.
• The idea of local foods being used in emergencies could be expanded to include other local foods for beneficiaries, for example, cereals and pulses for a general food distribution. Prior to this, a full market analysis should be undertaken. FFVs complement the Productive Safety Net Programme as they target beneficiaries who are most vulnerable to adverse nutrition impacts. An extended pilot in this and one additional woreda during the following months would further help to test the viability of this model and performance of CMAM indicators (particularly TSFP indicators).
• A monitoring and evaluation plan needs to be built into the design of the next FFVP. A baseline survey that includes assessment of dietary diversity, meal frequency and anthropometry would provide more objective evidence of impact. Ideally, a control area should be included in the evaluation. If TSFP performance indicators will be used for evaluation purposes, comparison woredas will need to be selected and any potential confounding variables assessed adequately.

For more information, contact: Pankaj Kumar, email: pankaj.kumar@concern.net

---

¹Ethiopia is divided into regions, which are further divided into zones, which in turn are divided into woredas. Kebeles are small village level administrative areas.

²The standard ration is 1 litre of vegetable oil per beneficiary per month and 6.25 kg of CSB/Famix per beneficiary per month, premixed before distribution.

³The seven food groups per international guidelines consisted of cereals, grains, roots and tubers; pulses, legumes and nuts; milk and dairy products; meat & poultry; eggs; vitamin-A rich fruits and vegetables; and other vegetables and fruits. Oils and fats were not included as a food group.