Targeting Emergency Food Aid - Experiences in Ethiopia

The Relief and Rehabilitation Network(RRN), has recently published a network paper entitled 'Between Relief and Development: targeting food aid for disaster prevention in Ehtiopia'. This paper focuses on the question of how food aid can best be targeted to the neediest household in food-insecure areas. This is looked at in the context of the 1993 National Policy on Disaster Prevention and Management (NPDPM) and its central strategy of channelling relief food through employment generation schemes (EGS) in place of general free distributions. In addition to contributing to the development of infrastructure the requirement to work for relief food (or for cash where available) is expected to discourage the less needy from seeking aid, i.e. to function as a targeting mechanism.

The debate about providing emergency food aid through employment provision centres around the choice between self (e.g. community targeting mechanisms) and administrative targeting (e.g. selecting beneficiaries on the basis of data collected by an external agency). The author asserts that self/community targeting has received little attention in the literature. A review of previous Ethiopian experiences with targeting through public works shows little evidence that it successfully self-selects the poorest and excludes the relatively better off even at low payment rates.

Two conditions essential for self-targeting are outlined ; an appropriate wage (in terms of food) policy and ability of the scheme to employ those who want work . However, in Ethiopia there are problems with both, particularly in crisis situations. Unlimited job provision on the scale needed to cope with Ethiopia's recurrent food security problems, may not be possible (given resource and management constraints) or desirable (given the government and donor policy objectives of limiting aid).Also the strategy of setting low wage rates with the objective of targeting the poorest may backfire as it may be only the slightly better off who can afford to have a hoousehold member working for below subsistence wages. Poorer families who are typically smaller require all manpower for subsistence activities. The little evidence that is available indicates that the relatively rich often volunteer as readily as the poor even at low payment rates. Examples of this happening are given from WFP, IFPRI and SOS Sahel employment creation projects.

A summary of views expressed by beneficiaries and implementing staff in areas receiving food aid sheds further doubt on the potential of pure self-targeting to meet the objectives of the NPDPM. It also suggests that this form of targeting is not easy or cheap.
Within Communities there seems to be a strong preference for sharing aid as widely as possible - this applies to employment entitlements.
In almost all cases encountered, household level targeting was administered by community representatives or committees of various kinds under the authority of Peasant Associations There was a remarkable level of participatory decision making with great efforts at fair targeting. However, community leaders found it an extremely burdensome task which tended to generate conflict. Everywhere it was maintained that all the people were poor and it was difficult to exclude some. A widespread opinion was that targeting invariably creates conflict but if there must be selection it is best done by the people who should gather together and select 'just elders' who can in turn identify the poor. Each village should select one or two elders. Four key elements for successful community targeting were identified:

• transparency of decision-making
• free availability of information to community members
• accountability of decision makers to constitutents
• an outside auditing or monitoring authority to detect corruption and support decision-makers when they have to make unpopular decisions.

The paper drew the following important conclusions:
i) The self targeting elements of EGS will not alone achieve the targeting objectives of the NDPDM. Given the scarcity of other employment opportunities there is always likely to be a high demand for EGS/FFW participation for all income groups.
ii) A combination of self-targeting elements with some additional system for household screening is needed. The realistic mechanism for this is not a highly administered selection system with centrally determined criteria and costly information requirements, but development of existing community structures for prioritising the poorest. However, community targeting is not a cheap or easy option. The costs to the community decision-makers themselves, in time and trouble are also considerable.

A final conclusion of the author is that area level targeting is is an essential first stage in the distribution system for countries like Ethiopia and is probably where the greatest potential gains in effectiveness and efficiency can be made.

Ref: Between Relief and Development: targeting food aid for disaster prevention in Ethiopia, Kay Sharpe: RRN Network Paper No 27, September 1998.