Community Managed Targeting of Emergency Food Aid: Does it Ever Work?

Alternative Distribution Systems: Tanzania - Food distribution in the Lumasi camp.

Malcolm Rideot is the SCF Country Programme Director for Tanzania. He has been instrumental in establishing the community managed targeting programmes which is the focus of this article.

As food aid budgets shrink, the concern to make sure that food aid is utilised by those who most need it is a legitimate one, and there is much ink spilt in trying to make sure that there is a rational, demonstrable and repeatable way of targeting food to the needy and distributing it in an equitable way. These targeting systems tend to have as their focus the need to account to the providers of the food or decision makers in Government, rather than ensuring an involvement with the communities served. This is the case despite the stated aim of most aid projects to be inclusive and responsive to communities, as is set out in the Red Cross Code of Conduct - now signed up to by many agencies.

It is interesting to contrast the provision of food aid to food insecure populations in rural Africa with the administration of Social Security benefits in a developed northern country. In the UK (the system with which the writer is most familiar) a huge amount of effort is taken by the state to prescribe who is entitled to what. However, even with the massive resources and levels of information available in the UK, there is still a significant amount of fraud, and politicians continue to make political capital out of introducing yet more levels of control into the system (vouchers instead of cash for example).

In food distributions in rural Africa, the distributing agency will often have no power over an individual (except perhaps the withholding of benefits) and relatively little information on the people in the community served. Agencies are still under the same pressure to control and make sure that the scarce resources devoted to food aid are used well, and by those who are most in need. The (usually unspoken) model by which these actions are judged come from social security models, which are not appropriate given the reality of emergency food distributions in developing countries. Distributing agencies have: no power, knowledge that essentially comes from outside the community, limited time and a set of assumptions about equity and entitlement. These assumptions are often driven by donor policies set thousands of miles away.

From the perspective of potential food aid recipients, the whole process places the recipient in an uncomfortable position, as each person has to prove that their hunger is greater than their neighbours hunger, according to criteria set by outsiders using rules that have not been discussed and may be poorly understood. Furthermore, as the resources are seen as being provided from outside the community, there is no feeling of ownership and the resources are scrambled for by all, a situation where there is a real danger that the poor and most needy will lose out.

Given that food aid decisions are made by donors and governments remote from affected areas, the ability of food distributions to be inclusive and centred on the recipients of the aid is always going to be difficult. People are subject to outside assessments and decisions about assistance. It is possible to begin to get away from the priorities of donors and their procedures and more towards the priorities of affected people by devolving as much decision making as is practical. This is an idea that is sometimes surprisingly hard to sell, even with the near universal commitment to people-centred development. In the example below handing over power has been shown to have advantages not only in the efficiency and timeliness of distributions, but also in their accuracy and equity.

Community Managed Targeting in Tanzania

In Tanzania, Save the Children Fund and the Tanganyika Christian Refugee Services (TCRS) have used a community based approach to try and place as much power as possible in the hand of the communities in a food distribution in Singida Region. The distribution made 13,000MT available to about 450,000 people between October 1998 and April 1999.

Community managed distributions are not a new idea, and various ways of carrying them out have been used. The basic idea behind them is that the people who receive the food should understand the targeting criteria being applied, have a say in them, know how much food is coming, and have a direct appeal and some control over decision makers.

In Singida the decisions over who was to get the food and its subsequent distribution was entirely in the hands of elected members of the community. Furthermore, distributions would not be given out on a month by month basis, forcing people to be on hand every month but be in as much bulk as was practicable so that people could migrate, pay debts or whatever was the best option for them.

The methodology first had to be sold to WFP and the government, as there were concerns expressed that if the community was left to get on with the distributions that they would be hijacked by the powerful, that the 'needy' would not get the food and that the bulk distributions would encourage sale rather than consumption of the rations.

Once it had been agreed with WFP and central government, the methodology was introduced to the regional government. The strength of the method was immediately seen, and the Regional commissioner gave instructions that the people in charge of the distribution should not be ex officio village government members but an elected committee for the purpose. This was an important step, as the whole methodology relies on the communities having confidence in, and control over the distribution process.

The decision over which areas were to get the food were decided on earlier assessments of crop failures and livestock numbers. This led to the production of a list of 'most affected' villages.

Alternative Distribution Systems: Tanzania - Benaco Camp, Ngara.

During the assessments, wealth ranking exercises had given the impression that about 60% of the population in the villages that were targeted were most in need of relief food and so initially 87 villages were targeted to receive rations for five months, a first round of a three month ration and a second round for two months. In the event, ongoing assessments of vulnerability and harvest prospects led to the inclusion of more villages and a sliding scale of ration provision based on anticipated harvest date. In the end, a total of 187 villages received food aid.

Stages of Implementation

The first stage of the community managed targeting process was to send a message through the local government system giving notice of a public meeting to be held to discuss the provision of food aid. At the meeting, facilitators gave notice that food was to come to the village, that the amount was limited and that it was for distribution to poor people as there was not going to be enough to go round. The decisions as to who would get the food and the distribution was to be left to an elected committee, which had to have a man and a woman from each sub-village on it. The meetings were well attended and it is estimated that at least one person from each household came. The election of the committee was by affirmation or a show of hands and most often people other than the village government officers were chosen. The committee then met with the facilitators, who discussed in more detail the wealth breakdown of the village and got the committee to come up with criteria that described the poor of the village. These criteria varied from place to place. The agreed criteria were then presented to the meeting and a time set for a second meeting, to take place once the committee had registered those to whom the criteria applied.

A second public meeting was held at a later date (depending on the time taken to register the villagers by the committee ) at which the list of those who were to receive food was read out and approved by the village. The amount of food to be delivered was again publicly confirmed and a date set for delivery. The weight of standard scoops (provided) was also publicly discussed.

The public meetings included the involvement of SCF and TCRS in the process. The food was consigned direct to the village from the store and signed for by the village committee It was then distributed as per the agreed list.

There Were Some Difficulties

The system is deceptively simple but tying together the need for public meetings, varying times for registration in villages and the logistics of dispatch of food is a major organisational headache, and it was important to keep the villagers and their committees informed of any changes as they are an integral part of the system.

The system was much appreciated by the recipients of food who saw it as a fair and dignified system which was understood and accessible. There were problems, as many of the population figures were based on centrally-held statistics that sometimes turned out to be wildly inaccurate. This necessitated the delivery of extra food to some places. In a village where the population estimate had been too high, the village committee handed back to SCF the excess for re-allocation.

The wealth breakdown in the villages gave the committees a hard job in registration as tough decisions had to be made. There was some flexibility in the proportion of the people to receive food, but it was made clear that simply spreading the food to everybody would mean that no one got any meaningful rations.

The committees by and large did a good job, as was expected, because they were chosen by their neighbours and knew the people involved well. There was one case of a committee holding back food, but this was sorted out by the villagers who promptly arrested the committee and made them return the food. In another instance the committee had to ask people who had received food to give a proportion back as the committee had got the arithmetic wrong. This was done by the recipients without demur as they had a clear understanding of the mechanics of the distribution and the entitlements of recipients.

The Programme Worked

SCF and TCRS were keen to learn how the distributions could be refined, and also what happened to the food. There was therefore a programme of ongoing assessment and distribution monitoring.

Alternative Distribution Systems: Tanzania - Rwandan refugees returning to Rwanda, lined up for rations of biscuits.

The 'food basket' monitoring was aimed at finding out who had received the food and what they did with it. A sample of villages were monitored by visits after the distributions had taken place. It was found that about 62% of the population received food (target was 60%) and that most of the food was consumed in the home, although a significant proportion (7%) was served to 'guests'. The people who received food were the poorer members of the village. The tightness and accuracy of the targeting exceeded expectations, although this is perhaps more a reflection of a lack of faith on the part of SCF and TCRS in the abilities of communities to organise themselves effectively more than anything else.

The programme was judged a success most importantly by the people who participated in the programme and it was seen to be fair, honest and understandable.

The distributions were also cost effective, as there was no need to engage large numbers of people to carry out registrations and distributions. It was fast, as once villages had set up their committees and knew what to do, any number of villages could simultaneously register people and carry out distributions.

The programme worked well, and it is interesting to speculate if it would work as well in a place where social relationships are not as strong (e.g. in a refugee camp). The problem with the system is more with agencies and donors rather than the recipients, as the system does not mesh well with some reporting requirements. It is impossible to provide a list of beneficiaries until all food has been distributed and the type and number of people that will be served can only be estimated in advance. The agency managing the distribution cannot account for the food to the level of recipient, as the responsibility for distribution stops at the village.

Donors, UN and NGOs alike all now share the language of inclusiveness and community involvement. If this is a real commitment it implies a letting go of power to the people that are served, with no holding back. The method above relies on the fact that people had information on which to decide and were very clear that decisions on their resources were made within their community. The dialogue and devolution was as real as it could be. Why should we be surprised that communities manage to organise themselves more fairly and with less fuss than outside agencies when they are given the opportunity?