In 2019 DevTech undertook research for USAID concerning food assistance in the Somali region of Ethiopia. The study had three objectives:(1) to assess whether the variety of Food for Peace commodities was appropriate, based in part on recipient food preferences;(2) to evaluate the potential interest and ability of food processors in transforming in-kind staple food commodities to meet large scale humanitarian assistance; and (3) to suggest the way forward for humanitarian assistance based on the study results and in light of political and security developments in the region which borders Somalia to the north and east and Kenya to the south. Background research revealed that USAID had not previously investigated recipient commodity preference, despite decades of humanitarian food assistance across the globe. DevTech, therefore, engaged several academic experts, who had been involved in related research, to serve as part of the study team.

The study found that while dietary patterns and thus nutrient intake in the region were very poor, most food assistance recipients were unsatisfied with at least portions of the commodities provided, cereal (red sorghum and maize), pulses, and cooking oil. While preferences varied in sub-regions, the main food item, red sorghum, was particularly disliked due to poor taste, which the study linked to the distribution in the south of the region of poor-quality cereal and pulses (e.g., spoilage, insect infestation). Recipients throughout the Somali region were widely satisfied with the processed cooking oil. While households consumed the oil, a large proportion of the recipients reported that they did not consume the rations of sorghum and pulses. What food that was not eaten by pests was generally sold to traders to be consumed in other regions of Ethiopia where preferences differed.

In terms of the second study objective findings included limited food processing in the region (mainly wheat flour) but interest on the part of enterprises, as well as underutilized manufacturing capacity to process commodities for use in humanitarian programs such as USAID’s. However, all the small and medium enterprises had questions and uncertainties regarding raw materials, the availability and/or cost of operating credit, and infrastructure constraints (electricity and roads).

The final study objective concerned future food assistance modalities, i.e., in-kind food, vouchers, or cash. Data collected through the study’s large-scale, representative survey revealed that most current USAID food aid recipients preferred cash assistance. Key informant interviews with government and donor representatives found a preference for cash, as it strengthens both regional and local markets including the participation of food processors, traders, and local vendors. Most all of those in favor of cash assistance noted, however, that it was not the best modality in all circumstances (places). Interestingly, non-beneficiaries surveyed preferred in-kind food over cash. Researchers speculated this was because they had not experienced the poor taste/quality of the distributed commodities noted by the majority of surveyed recipients. In only one isolated area was the use of vouchers preferred, presumable because the local economy was largely non-cash based.

The best food assistance modality has been debated among humanitarian aid specialists and economists for decades. This long-standing controversy has been joined by the more recent debate concerning the use of cash transfers, conditional or not, in lieu of non-cash assistance in a variety of development sectors. One important feature of cash assistance programming for whatever purpose is that it is typically the most cost-effective means of delivering assistance, particularly when cellphone-based money transfer is utilized – as the study examined. Research indicates that if the overall objective is to improve welfare, then cash assistance is preferable, although vouchers appear to do better when increasing dietary diversity is considered.

Recent studies of all types of cash transfer programs have been largely positive, although they do note that such programs can have “negative spillovers,” at least on non-beneficiaries, if not designed carefully. World Bank research found, for example, that in Malawi the Social Action Fund program cash transfer resulted in local food commodity prices going up in smaller, more isolated local economies as a result of the influx of cash to beneficiaries. As a result, food insecurity among non-beneficiaries was recorded as they struggled to buy higher-priced food. DevTech’s study warned of a similar consequence of the use of cash assistance in certain areas of the Somali region poorly connected to markets and recommended careful assessment of local food availability before ramping up of cash assistance programming. When introducing cash assistance time may be needed for market-based supply to catch up with the increased demand.

Despite the possibility of any such negative spillover, the DevTech study concluded that in much of the Somali region cash was the best food assistance modality and thus the way forward for humanitarian assistance in the region. Exceptions to this modality were dependent on the availability of food in local markets (and thus price sensitivity), as well as other factors such as the type of household (e.g., pastoral vs. semi-urban) and seasonality. In short, no food assistance modality solution fits all. This study concluded that consumer preferences are important to consider, as is careful, geographically specific investigation to discern local market conditions in advance of selecting food assistance modalities in the Somali region of Ethiopia. This conclusion appears to be relevant elsewhere in the world where the U.S. Government and others provide this form of humanitarian assistance.


  1. USAID/Ethiopia. Somali Regional Consumer Commodity Preference Study. DevTech, February 2020.

  2. World Bank. In Fighting Poverty, Cash Transfer Programs Should Be Wary of Negative Spillovers. On-line Feature Story, December 27, 2019.

  3. Hidrobo, Melissa, et. al. Cash, food or vouchers? Evidence from a randomized experiment in northern Ecuador. Journal of Development Economics 107 (2014).