Fresh food vouchers: findings of a meta-evaluation of five fresh food voucher programmes

Action contre la Faim (ACF) has been implementing cash-based interventions since the late 1990s. ACF’s main focus is the treatment and prevention of acute malnutrition, and it has used cash-based responses to pursue this objective, including using vouchers to increase access to fresh foods (vegetables, fruit, eggs, meat, milk and fish). Fresh food vouchers (FFV) have provided households with complete food baskets, or have been used to supplement staple foods with fresh micronutrient-rich foods. FFV programmes also support local markets and traders. Fresh food vouchers can be used in slow-onset as well as acute crises to provide short- or longer-term support to people highly vulnerable to acute malnutrition.

ACF commissioned a meta-evaluation of five of its FFV programmes in Bolivia, Kenya, Haiti, Pakistan and the occupied Palestinian territory (oPt). The evaluation was funded by the Cash Learning Partnership (CaLP) and ECHO, and the results were used as the basis for the development of a good practice guide on FFV. ACF, Emerging Good Practice on Fresh Food Vouchers, 2011.  The OECD/DAC criteria (appropriateness, connectedness, coherence, coverage, efficiency, effectiveness and impact) were used as the basis for the evaluation. The methodology consisted of a review of the independent final evaluations of each of the voucher programmes and other related documentation, and interviews with headquarters and field staff.

Assessment and causal analysis

In each intervention, ACF undertook needs assessments and causal analyses of malnutrition, but with different levels of rigour and detail. Some interventions established links between underlying causes and their contribution to malnutrition, while others used a household dietary diversity survey to identify gaps in household and individual food consumption. All needs assessments found poor dietary diversity, largely caused by lack of access to fresh foods. Causal analysis of micronutrient malnutrition was not based on a detailed consumption recall, nutrition value survey and metabolic test, with the exception of Bolivia. Table 1 reviews the needs assessments carried out in the five FFV programmes.

Objectives

The objectives of the FFV programmes included increasing dietary diversity and ensuring an adequate diet, reducing micronutrient malnutrition, preventing mortality, preventing malnutrition and reducing negative coping strategies. All five programmes used paper vouchers which could be exchanged for fresh food in local markets; in Pakistan the vouchers could also be used to purchase staple foods. In Bolivia, Kenya (Dadaab refugee camps) and Haiti, the FFV was designed to complement general food distributions and supplementary feeding programmes, providing fresh foods like vegetables, fruit, eggs, meat, milk and fish. In the flood response in Pakistan in 2010 and 2011, vouchers replaced ACF’s general food distribution once sufficient staple foods were available on the market. In oPt, the vouchers were initially targeted at families vulnerable to food insecurity, but who had sufficient resources to cover their staple food needs. Programmes in Bolivia, Kenya and Haiti were coupled with nutrition and health education sessions.

Results and cost-efficiency

FFV increased the dietary diversity of households in all programmes, but to different degrees. This disparity resulted from weaknesses in programme design and implementation that could be better managed in the future. In the case of Haiti, dietary diversity only increased marginally compared to pre-earthquake levels, which were already poor. This was in part because households did not receive general food rations – meaning that households received FFVs when they actually needed staple food products (it was assumed that households would receive World Food Programme (WFP) general rations, but this did not happen due to a change in government policy). The value and selection of earmarked food items redeemable with the vouchers were defined at the design stage of the programme (see Table 2), Voucher items were permitted in the order they were most frequently purchased: cereals, flour, pulses, legumes, nuts, sugar, honey, tea/coffee, vegetables and spices, oil, meat, fresh/dried fish and sea foods, milk and milk products, biscuits, eggs, salt, fresh/dried fruit and juices. and were sensitive to changes outside of the ACF projects, and hence were in the end not always sufficient to achieve the overall programme objectives.

The evaluations found other possible impacts of FFV, although some cannot be directly attributed to the FFV programme given other potential contributory factors:

• Increased haemoglobin rates in Bolivia were attributed to increased consumption of iron-rich foods (fresh vegetables and meat).
• In Dadaab, nutritional programme attendance increased due to large-scale camp mobilisation and people’s interest in having their children screened for malnutrition. In Dadaab and Haiti acute malnutrition rates declined, but this cannot be attributed to the FFV alone. FFV vouchers contributed to improving the quantity and quality of food, which in these two contexts were associated with underlying causes of acute malnutrition.
• Reallocation of income to other livelihood needs and hence the protection of assets (oPt and Pakistan) – the FFV supplemented households’ food expenditure, enabling people to invest the income saved in livelihood recovery.

As none of the programmes systematically measured participants’ knowledge, attitudes and practices (KAP) in relation to nutrition, it is difficult to know whether encouraging people to eat fresh food will have a lasting impact on their diets. Without this baseline information, it is also not clear whether the simple provision of cash or vouchers, instead of the package of vouchers and nutrition education provided, could have achieved the same impact.

Only the Bolivia evaluation attempted a rigorous cost comparison. It concluded that a voucher, even where market access is limited and significant transport costs are incurred by beneficiaries, was 15% cheaper than delivering in-kind food rations.

Monitoring FFV programmes

Establishing baseline data and monitoring FFV programmes was a key challenge. Both process and impact monitoring indicators need to better reflect the programme objectives, and need to include indicators of a healthy diet, particularly when the outcome indicator is the child’s nutritional status. While household dietary diversity and food consumption scores support the monitoring of overall household consumption, individual dietary diversity (children under five) provides more insight into the impact of FFV on nutrition objectives, as in the Kenya and Haiti projects (Table 3). The individual dietary diversity score for children under five years of age used eight food groups. The household dietary diversity score used 12.   4.9 was the baseline after the floods and in the absence of assistance. With the GFD the HDDS went from 4.9 to 7.7. After the GFD finished and the FFV programme was implemented (which included staple foods) the HDDS went up to 9.0. The figure 4.9 is retained here as a baseline in order to compare the relative increase during the GFD and FFV phases.  It is most important to monitor nutrition objectives for under-fives, as they are most susceptible to undernutrition. Improvements in individual (children under five) rather than household consumption patterns are more relevant to assessing the potential impact on undernutrition prevalence and prevention.

Markets

Where market assessments were rigorous (in Haiti, Pakistan and oPt) and market monitoring systems were applied, interventions monitored supplies, adjusted the value of the voucher in line with changes in local prices and assessed the impact of the programme on local markets. In all programmes, fresh food vouchers supported markets and increased the incomes of participating vendors, particularly benefiting women vendors in Haiti and Dadaab, shopkeepers in Pakistan who were affected by floods and the dairy sector in oPt. In Dadaab, the increase in demand for fresh foods (as a result of the programme) increased the supply of certain fresh foods that had not been previously available in the camp, therefore benefiting other camp residents as well.

Emerging good practice in fresh food vouchers

The meta-evaluation has identified emerging good practice in FFV programming. Key elements include: • A good fresh food voucher project starts with an adequate needs assessment that includes an assessment of food consumption and the likely causes of poor diet, including knowledge, attitudes and practices. Fresh food market assessments are essential to design interventions to support markets and to gauge the capacity of markets to meet needs. Where market recovery is an objective and market assistance is provided, vouchers can help to increase demand.

• If malnutrition is an actual or potential problem and feeding, care and health practices may be contributing factors, assessments must include the individual child’s and household’s food consumption, as well as infant and young child feeding practices.
• Increased consumption of micronutrient-rich fresh foods can address micronutrient deficiencies (MND) if consumption is the cause (as opposed to absorption or utilisation). The causes of deficiencies must be identified and understood in order to design a voucher and nutrition education programme that directly addresses the relevant nutrients and behaviours.
• The design of voucher programmes and complementary work, such as public health promotion, should reflect the nutritional analysis and the specific food requirements of the target group, and promote the best use of the vouchers. If staple foods are an assessed need, the voucher should include staples, or a reliable source of staple foods must be guaranteed.
• Food vouchers (and other cash-based interventions) can be used as an incentive to participate in health and nutrition programmes, increasing the coverage and possibly the effectiveness of these interventions. However, care must be taken in design and monitoring to ensure that there are no unintended negative effects, for instance intentionally starving children to make them eligible as beneficiaries. Voucher programmes that aim to improve micronutrient consumption and reduce micronutrient deficiencies, such as anaemia, require robust monitoring systems.
• The importance of adequate planning, staffing and monitoring and financial systems that ensure prompt payment of participating vendors has been highlighted, though this is not specific to FFV programmes. With experience and advances in new technologies these areas are improving. Accountability systems should include participant feedback, and programmes should be responsive to participant and vendor needs.
• Effective strategies and systems must be put in place to ensure that all stakeholders (the humanitarian agency itself, participants, vendors and the broader community) promote the intended use of the voucher (compliance) and reduce the likelihood of counterfeiting.

The meta-evaluation established that fresh food vouchers can be effective in improving food consumption in emergencies as a complement to general food distributions, when households have sufficient access to staple foods and/or when the voucher includes staple foods. Fresh food vouchers also show potential as a complementary intervention to improve the impact and outcomes of other nutrition-related activities, such as supplementary and therapeutic feeding and nutrition education. Close coordination between teams and departments is necessary to ensure coherence between approaches. Of critical importance to FFV interventions is improving nutritional causal analysis to define clear objectives and indicators. New technologies related to voucher design and transfer mechanisms, such as electronic vouchers, will reduce the administrative workload. More experience with and learning from FFV programmes will provide further insight into their appropriateness and cost-effectiveness in preventing, reducing or treating acute and micronutrient malnutrition, either alone or in combination with other interventions.

Silke Pietzsch is Senior Food Security and Livelihoods Advisor, ACF USA. Muriel Calo is Food Security and Livelihoods Advisor, ACF USA, Julien Jacob is Food Security and Livelihoods Advisor, ACF Spain, and Julien Morel is Food Assistance and Social Protection Advisor, ACF France. The authors would like to thank Maria Bernardez (formerly ACF Spain) for contributions to the meta-evaluation process, and Kerren Hedlund, who conducted the evaluation.