Food Aid: What we know and what we need to know

UNU-WIDER / Aug 2012

Women carry their ration of food. © UN Photo/Tim Mc KulkaIn the 1960s food aid made up nearly 20 per cent of overall overseas development aid, today that figure is five per cent. Increasingly food aid is provided as emergency relief rather than as a form of long term support and many donors have expressed strong interest in switching from food to cash assistance.  Food aid continues to generate heated debate, particularly around what effect it has on food consumption and nutrition, food markets, and labour supply, as well as the extent to which it exacerbates or mitigates conflict. In the UNU-WIDER working paper 'Mapping the Impacts of Food Aid: Current Knowledge and Future Directions', Amy Margolies and John Hoddinot provide an overview of what we currently know about the impact food aid has on these issues and highlight key areas in which further research is needed.

The impact of food aid

The most common way by which food aid is distributed to a beneficiary population is through a 'standard food ration', which typically consists of a basket of dry or canned food items being shipped to recipients by international donors. Food aid is also delivered through school meals, in return for labour and through the distributing food purchased locally. There is a diverse range of literature discussing the effects such aid has on recipient households and agricultural practices. A number of interesting findings can be drawn from these discussions.

Margolies and Hoddinot point out that food aid transfers do seem in general to increase household calorie consumption.  The evidence for this comes from two countries where food transfer programmes have been studied in detail, Bangladesh and Ethiopia. In Bangladesh the amount of this increase differed from programme to programme. The highest increase recorded was associated with the transfer of micro-nutrient fortified whole wheat flower (66 kcal per taka), the lowest with a program that offered participants rice in return for participating in a public works projects (23kcal per taka). In Ethiopia studies show that food transfers put into place after the drought in 2002 increased both total consumption and pre-school nutritional status.

One particular concern of many critics of food transfers is that they will disincentives recipients from participating in the labour market and acts as a substitute for local food production. However a number of studies show that this is not the case. For example, an analysis of data from rural Ethiopia revealed no evidence of negative dependency effects, and, in fact, showed a potential positive effect on labour supply. One might fear disincentives to agricultural production if food transfers provide an adequate replacement. However, in a strange case of aid inefficiency having a positive unintended consequence, due to the inconsistent timing of the deliveries associated with food transfer programs food aid does not seem to lead to recipients not participating in the labour market.

Some critics of food aid argue that it leads to lower food prices. The increased supplies are thought push down prices, and thus disincentivize agricultural production in the area. Early studies into the impact of food aid provided evidence consistent with this hypothesis, however Margolies and Hoddinot point out that empirical work done over the last fifteen years has largely failed to replicate these adverse results. In particular a study of the effect of food aid, in the form of yellow maize, in Mozambique found that the program had no noticeable effect on prices and studies of programs of Kenya and Swaziland have replicated these results. While some studies have found that food aid has had this effect they have put this down to programmatic failure rather than problems with food aid itself.

Emergency Food Aid

Food aid is becoming increasingly restricted to emergency settings that arise often as a result of internal strife in recipient countries. There are particular considerations that need to be taken into account by both practitioners and policy makers when thinking about food aid provided to countries experiencing these circumstances. In particular it is important to question whether food aid is likely to exacerbate or mitigate any ongoing conflict.

Margolies and Hoddinot suggest that if the conflict is driven by grievances about the distribution of resources or by considerations of survival then food aid does have the potential to serve as a mitigating force. The disbursal of food aid to regions which feel they have been deprived is a mechanism by which the inequalities at the heart of the conflict can be diminished. In the case of conflicts driven by considerations of survival, if food aid is delivered at a level sufficient to meet the survival thresholds of both sides then it can clearly reduce the incentives for conflict.

However there are a number of cases where it appears that food aid has prolonged conflicts. There is evidence that during the Rwandan genocide the government redistributed food aid to elites, thus further exacerbating tensions and perpetuating inequalities. Similarly in Sudan and Ethiopia there is evidence that food aid was diverted by the governments to supportive constituencies and the military. While there may be a theoretical basis for believing food aid has the potential to mitigate conflict, in practice there are many accounts of its failure to do so and few, if any, documented successes. Indeed, Margolies and Hoddinot point out that one of the few studies focusing specifically on the relationship between food aid and conflict found that increasing food aid has the effect of increasing the incidences of conflict, an effect which is even larger in countries without a civilian government. There is evidence that other forms of aid aimed at community and peace building can in some circumstances help reduce the incidence and duration of conflicts. Given the increasing tendency for food aid to be restricted to emergency settings, more work clearly needs to be done in this area.

Alternative approaches: food, cash or something else?

Elderly woman receives emergency food aid. © UN Photo/Tim McKulka

On the face of it there seems to be a number of reasons why both donors and recipients might prefer cash transfers to food. Cash transfers are in theory more economically efficient because they allow beneficiaries to buy what they need most and thus have a less distorting effect on production and consumption choices. Cash may also stimulate agricultural production by raising demand in the local market. It is also much more expensive to distribute food rather than cash as administrative costs which are 20 to 25 per cent higher. Furthermore a study of the effects of food aid in the aftermath of the Asian Tsunami of 2005 found that recipients of cash, rather than food, transfers were more likely to improve the variety of their diets. Given the supposed benefits of cash transfers it is important to consider whether there is any reason why they should not replace all food aid.


One potential reason is that cash transfers could be ineffective under certain local conditions. If recipients have no physical access to markets then having money to spend will do them little good. Furthermore, if traders lack the capability to bring extra food to the market without significantly raising prices then a large portion of a cash transfer will be eaten up by inflation. Cash transfers to areas with a very uncompetitive market may simply allow traders to raise prices.

It is also the case that many recipients simply prefer food to cash. This preference is obviously dependent on prevailing food prices and tends to increase as prices rise, however recipients also claim local traders, aware of the cash transfers, collude to raise food prices to levels they know their customers can newly afford. Margolies and Hoddinot argue that we should take it as a general principle that recipients should have some say in the form of the assistance they receive, given this, the preference food is clearly an argument in favour of that kind of transfer.

There are a number of factors that should effect the decision about whether cash or food transfers are preferable and many of them are not yet fully understood. Further research in this area is particularly important given the growing gulf between donors (who prefer cash transfers) and recipients (who prefer food).

The authors conclude by pointing out that while programme designers need to be mindful of the potential negative effects of food aid, there is little evidence that it disincentives its recipients to work or that it has adverse effects on food markets. However with regard to the impact different food aid modalities have at a household level and during times to crisis the author argue that more work needs to be done.

This report by James Stewart summarizes UNU-WIDER working paper no. 2012/34, 'Mapping the Impacts of Food Aid: Current Knowledge and Future Directions' by Amy Margolies and John Hoddinott.

Related article: How can food aid be more effective?