Does food aid support or undermine livelihoods? (Special Supplement 3)

4.1 Introduction

A WFP ship carrying food aid docks in Indonesia

Food aid remains the over-riding response to emergencies, regularly constituting over half of consolidated emergency appeals (Development Initiatives, 2003). Food aid can take many forms (see box 6) and plays an essential role in saving lives in many emergencies, and supporting livelihoods in some. However, there are many emergency situations where food aid is not necessarily the right response to address food insecurity or the impact of disasters on livelihoods.

"General food distribution may not be appropriate when:

• adequate supplies of food are available in the area (and the need is to address obstacles to access)
• a localized lack of food availability can be addressed by support of market systems."

(The Sphere Project (2004). Guidance note 1 under food security standard 1).

Both donors and beneficiaries have also used food aid as an economic resource, since it releases income (otherwise spent on food) which can support livelihood strategies and help build up assets. The impact of food aid on production based strategies is more controversial and there is evidence of both negative and positive impacts. Finally, food aid policies and institutions at national and international level have critical impacts on livelihoods. At national level, for example, the use of food aid in strategic grain reserves as part of social safety nets can be an important and effective emergency response, strengthening the ability of poor or destitute populations to meet their basic needs. At international level, structures for global food aid governance, such as the Food Aid Convention (FAC) (FAC, 1999) and the World Trade Organisation (WTO), are essential for regulating the allocation of food aid according to need.

This section reviews briefly the use of food aid as support for livelihoods assets and strategies, and also examines some of the concerns around the potential disincentive effects of food aid. The possible disincentive effects of food aid, on trade, markets and production, have recently gained prominence in the negotiations on food aid as part of the agriculture agreements in the WTO¹¹.

4.2 Food aid as livelihood support

Children queuing to eat in the ACF supported canteen in Haiti (see box 7)

Food aid can be a form of livelihood support either when provided as general rations to assist in preserving or rebuilding assets, or as a FFW programme which creates community assets which promote livelihoods. The provision of food aid can indirectly contribute to livelihoods as a resource transfer. In the Haiti case example, (see box 7), where food aid is provided mainly for nutritional purposes, beneficiaries may view its main benefit as providing economic support.

As the food aid agency of the UN, the WFP also supports the use of food aid to support livelihoods, with the aim of preserving essential assets and assisting in recovery. Whilst the WFPs main priority is to use food aid to save lives, it recognises that more lives are saved in the longer term if people do not lose their livelihoods and become destitute as a result of a disaster. The WFP, however, faces serious challenges in achieving livelihoods objectives in emergencies. Emergencies pull staff towards those who are most in need, and when time and resources are insufficient (as is often the case), it may not be feasible to expand programmes to include those who also retain some assets (WFP, 2003, May).

WFP food being off loaded in Indonesia post-tsunami

Lack of resources has meant that food aid has rarely been distributed on a scale and of sufficient duration to prevent sale of, or recovery of, essential assets. In areas of chronic poverty, everyone in the affected population would need to be included. Furthermore, food distributions would need to start earlier in order to protect assets, and last longer than if the main purpose was only to save lives in order to promote recovery. This is particularly true for pastoral populations, which take longer to recover from drought (or other disasters where livestock are lost) than farmers or wage labourers.

The extent to which food aid by itself can really support livelihoods is unclear. Evaluations of the impact of food distribution on the pastoral economy have found that slaughter and sale of livestock has been halted and pastoralist purchasing power has increased due to increased livestock prices (Bush, 1995). An evaluation of the Oxfam Turkana programme in the early 90's considered the focus on saving livelihoods as well as lives to be justified (Wiles et al, 1993). Food distribution was also found to promote the maintenance of social networks through the sharing of food rations. Evaluations of programmes in both Turkana and Wajir have consistently found that sharing of food rations has been one of the main uses of food aid (Bush, 1995; Jaspars et al, 1997).

Source: ACF (France)

However, there have also been significant impediments to food aid having an impact on livelihoods in Sudan and Kenya. First, food aid has not been provided on the scale and for the duration needed for livelihood support. Second, droughts or other disasters have come in rapid succession thus not allowing sufficient time for recovery. Finally, and probably most importantly, food aid alone cannot achieve sustainable livelihoods or even livelihood protection or recovery. Other types of livelihood support, as well as policies at national level which promote the development of marginal groups, must be implemented in conjunction with food aid. The Oxfam programme in Turkana and Wajir subsequently started implementing a range of livelihood support interventions, including cash for work, re-stocking (during recovery), livestock offtake (at the early stage of emergency), and water interventions.

Source: Oxfam Port Sudan, Sudan

In FFW programmes, communities should benefit from the community assets created and from the food aid itself. Community assets to improve food security or livelihoods include infrastructure works, like road building to enable market access, constructing storage houses for crops, or digging wells to provide domestic water. Activities may also include water for livestock and fields, natural resource conservation programmes, such as soil and water conservation techniques or terracing. The WFP gives the following examples of food aid as livelihood support (WFP, 2003, May):

• WFP Afghanistan used food aid to help create urban and rural productive assets by supporting school feeding, women's literacy programmes, and urban bakeries.
• In the DRC, the WFP used food aid to rehabilitate rural and social infra-structure (for example, feeder roads, health and sanitation infra-structure), agricultural production (by distributing food along with seeds), and to encourage food insecure displaced or resettled women to attend vocational training.
• As part of flood response in Bolivia, food aid was used to recover flood damaged assets, such as small irrigation systems and agricultural land. In Colombia, the WFP used food aid to support construction of small roads and water and sewage channels, as well as vocational training.

A school feeding programme supported by WFP in Sri Lanka

There may be difficulties in implementing FFW programmes. Box 8 gives an example of the challenges faced and lessons learnt by Oxfam's food for recovery programme (FFR) in Red Sea State, Sudan, where food aid was required on a large scale for a widely dispersed population. Food for recovery is a less structured form of food-for-work were activities should contribute to initial recovery and should not require outside technical supervision (Sphere, 2004). In practice, the creation of community assets (whether as part of FFW or FFR) has often been secondary to the provision of food aid, and when this is the case, insufficient attention is given to community involvement in identifying projects, the capacity to maintain assets, or ensuring the necessary technical expertise is available to carry out the work.

4.3 Food aid, markets and production

Maize seller in Mozambique who said to Oxfam interpreter, "tell her not to bring any food aid here or no-one will buy my maize"

The potential negative impact of food aid on production and markets is of concern to many in the international aid community. It is feared that the provision of excessive levels of food aid will reduce the incentive for farmers to plant, and that lower market prices will reduce their income from the sale of crops, creating dependency on external assistance in the longer term.

There is strong evidence that food aid, in particular programme food aid, displaces commercial imports in recipient countries. All food aid will displace imports to some extent, as food aid will inevitably replace some portion of the food that recipients would otherwise purchase (Barrett and Maxwell, 2005). The potential for food aid to displace imports had led food aid to become a hotly debated topic at the WTO negotiations on the agriculture agreements, i.e. food aid is seen as a hidden export subsidy.

The degree of displacement, in turn, depends on how well targeted the food aid is, i.e. the additional amount consumed is likely to be higher for the poorer sections of a population. Two main forms of food aid, monetised food aid and programme food aid, are not targeted. The monetisation of food aid involves the sale of commodities in recipient countries, and the local currency is used to promote poverty reduction and food security initiatives. In 2002, about half of all project food aid channelled through NGOs was monetised (OECD, 2005). Programme food aid is provided as budget support, for example in the form of concessional sales. US programme food aid is specifically linked to the creation of overseas markets and almost all programme food aid is monetised. These forms of food aid have a much greater potential to have a negative impact on imports.

Women carrying food aid with visible donor logo

Evidence of impact on local markets and production is mixed. The impact of food aid on markets in the recipient country will also depend on the precision of targeting. Poor targeting of food aid, or mistimed deliveries, is likely to decrease the demand for local produce and decrease prices in local or national markets (Barrett and Maxwell, 2005). This reduces the income of local farmers. Although food prices will almost always fall, at least in the short term, the impact on farmers livelihoods depends on the balance between income lost through lower prices and the positive market impacts on inputs, capital and labour. For example, food aid could have a positive impact through relieving short term borrowing constraints, and the availability of food aid as a safety net can encourage producers to take more risks (Barrett and Maxwell, 2005).

Simple descriptive statistics can mistake a correlation between food aid receipts and low production for causality (Abdulai et al, 2004, June). Since food aid should be targeted at areas with low production potential, or with households who have low labour supply, we should expect this correlation. By controlling for other household characteristics, and by investigating the relationship between food aid deliveries and food production throughout Sub- Saharan Africa at the macro-level, food aid has been found, on balance, to be broadly stimulative to production. One possible explanation is that financial liquidity constraints so severely limit food production and distribution in much of Sub-Saharan Africa, that the ameliorative effect of food aid on farmers' access to inputs compensates for the downward pressure that food aid exerts on producer prices (Abdulai et al, 2004, June). The evidence at local level suggests that individual farmers do not base their planting decisions on the likelihood that food aid will depress prices. In most situations, the provision of food aid is too unreliable for farmers to make such calculations (Harvey and Lind, 2005, July).

Small stall in Thyolo district, Malawi where a cash transfer scheme is operating

In emergencies, it is possible for excessive levels of food aid to have disincentive effects. For example, in 2002 and 2003, food aid donors over-reacted to a projected 600,000 metric tonne (MT) food deficit in Malawi, and sent close to 600,000 MT of food in aid. However, commercial and informal importers brought in an additional 350,000-500,000 MT. As a consequence, maize prices dropped from $250 per tonne to $100 per tonne in the course of a year. Local production of maize, cassava, and rice fell and estimated losses to the Malawian economy were approximately $15m (Mousseau, 2004, March). Furthermore, the provision of food aid in kind can have negative impacts as it often arrives late, commodities are inappropriate, and quantities insufficient. In 2004, 74% of all food aid was sourced in donor countries (OECD, 2005), which may take up to 4-5 months to arrive in the recipient country.

The potential disincentive effects of food aid should be an argument for looking at the appropriateness of the assistance being provided, and the way it is provided, not whether it should be provided at all (Harvey and Lind, 2005). In countries or regions where food is available locally, it is both quicker and cheaper to purchase food aid locally. The cost of local purchase of food aid is, on average, 50% less than food aid sourced in donor countries (tied food aid) (OECD, 2005). Local purchase also provides commodities that people in need of food aid are used to, while local purchase can stimulate the local economy. A study of the statistical correlation between nations' annual per capita production of cereals with that of their neighbours, found that there was the potential for expanding triangular transactions, i.e. purchases within the region (Barrett and Maxwell, 2005).

4.4 Global food aid governance

The main global instrument for regulating food aid is the FAC¹². It is a legal agreement among seven donor countries and the EU. Recipient countries are not included. The latest FAC renegotiation (due in 2004) has been suspended until agreement could be reached within the WTO on food aid reforms.

The FAC was intended to enhance the capacity of the international community to respond to emergencies by guaranteeing a predictable flow of food aid every year. The 1999 FAC sets minimum aggregate commitments at 5.5 MT, whereas originally minimum commitments were set at 10 MT. Each donor has a minimum commitment per annum. The FAC also provides guidelines for the provision of food aid so that it is targeted effectively. For example, the 1999 FAC states that:

• Food aid should only be provided when it is the most effective and appropriate form of assistance.
• Food aid should be based on an evaluation of needs in the recipient country.
• Food aid provided is aimed particularly at the alleviation of poverty and hunger of the most vulnerable groups, and is consistent with agricultural development in those countries.

Carrying distributed food in Kenya

The FAC also encourages increased distribution of food aid through multi-lateral channels and the local or triangular purchase of food aid when possible. Members should provide regular reports to the Food Aid Committee of the FAC, on aid deliveries, arrangements for the future supply of food aid, and policies affecting the provision of food aid.

One problem with the FAC is that it is a voluntary code, with no compliance or legally binding regulatory mechanism. In practice, therefore, the guidelines are not followed, and food aid is not provided according to need. The committee on surplus disposal in FAO has a register of transactions which has a mechanism for notification of food aid shipments. It is also a forum for complaints about food aid issues. However, by 2001, reporting rates dropped to 4% (Barret and Maxwell, 2005). Akey issue is that current food aid regulatory mechanisms were established at a time when food aid was mainly a means of surplus disposal, and the main aim of regulatory mechanisms was to prevent commercial displacement. In recent years, with the growth of emergency food aid, food aid has become more demand led.

The WTO negotiations on agriculture, and the revision of the FAC that will follow, provide a unique opportunity to make global food aid governance more effective. The main issue will be how to make the FAC more binding and to ensure that a compliance mechanism exists. One way of doing this is by linking it with the WTO dispute regulation process. Barrett and Maxwell (2005, December) have given recommendations for a new Global Food Aid Compact (GFAC). The key elements of the Compact are:

• Inclusion of recipient as well as donor countries.
• Link with an international Code of Conduct where all signatories agree to role specific obligations.
• Commitment of donors to a minimum of financial contributions as well as minimum tonnages.
• Link with the WTO for its dispute resolution process.
• Establishment of a Global Food Aid Council, an interagency body with technical capacity to oversee and monitor the implementation of the GFAC.

The regulation of emergency food aid through WTO disciplines is a contentious issue. On the one hand, tied aid in emergencies is clearly not the most effective form of assistance in many emergencies and therefore its use should be regulated. On the other hand, any regulatory measures should not lead to the delay of emergency food aid shipments when they are urgently needed.

¹¹See WTO Negotiations on Improving Food Aid, By Suzanne Jaspars and Chris Leather, Oxfam GB, Field Exchange 26, p21. November 2005. Also, Oxfam International (2005, March).

¹²The FAC was originally agreed in 1967 as part of the International Grains agreement, and is housed in the International Grains Council in London. The FAC has been revised three times since its inception.