A refugee who arrives in Uganda today is under enormous stress.

From Seed to Shelter started researching Nakivale, Rwamwanja, and Kyaka II settlements knowing that food insecurity was a big issue, but the more people we spoke to in these refugee settlements, the more startled we became. The coping mechanisms that our interviewees had witnessed included selling house parts, pulling children out of school to work, early marriage, transactional sex, and suicide.

The gap that we found most problematic concerns new arrivals. Core to Uganda's refugee policy is the transition from support to self-reliance. With access to land, households are expected to rebuild livelihoods and, over time, rely less on humanitarian assistance. However, the pace at which refugees are expected to make this transition no longer maps onto reality. New arrivals arrive with minimal savings, without networks, and increasingly onto plots of land too small to feed a family. Food support now ends within 3 months of arrival, while a maize harvest takes 4. This gap is both predictable and, as far as we can tell, unaddressed.

Our team’s goal is to identify mechanisms for reducing vulnerability, and then direct partners towards ways in which they can support these. We will wrap up our research by December 2026 with a visit to the settlements, and publish a cause guide to set out where capital can make a difference. We welcome introductions to funders passionate about this issue who are keen to learn alongside us.

This year, FSTS was invited by Every Shelter to explore the food situation in Nakivale Settlement, Uganda. They had witnessed the impacts of humanitarian assistance reductions on the community they served, and suggested that this was a situation where we could make an impact.

We began researching the situation in January 2026, and expanded our scope to 3 settlements in south-western Uganda: Nakivale, Kyaka II, and Rwamwanja. To date, we have conducted 23 interviews with refugee-led organisations, researchers, and other stakeholders. In April 2026, we enlisted the help of a student group from the University of Oxford, From the Roots Up, to conduct a desk review of the academic literature around refugee food systems in south-western Uganda.

We are at a juncture where we have a reasonable sense of what we know, and what we don’t. We will conduct a visit to the settlements in December 2026 to complete our research, and turn our learnings into recommendations shortly after.

As a result, the World Food Programme (WFP) sharply reduced general food assistance in East African refugee settlements with a differentiated assistance model. Under this system, households assessed as least vulnerable were phased out of support entirely.⁹ In Uganda, the May prioritisation exercise reduced the number of food-assisted refugees from approximately 1.6 million to 662,000, a reduction of nearly 60%.¹⁰ Importantly, the recategorisation was driven by the funding shortfall rather than a comparable reduction in food need: many households were moved into lower categories, and some lost food assistance, without evidence that their underlying food insecurity had improved.

The impact on households has been worrying. The 2025 Food Security and Nutrition Assessment, covering refugees in Uganda’s settlements and Kampala, found that 65% lacked sufficient food, over 7 in 10 households reduced daily meals, and 1 in 4 pulled children from school.¹¹ IPC analysis projected that over 642,000 refugees (an estimated 32% of the population analysed) would experience crisis-level acute food insecurity or worse between July 2026 and January 2027.¹²

Hunger is year-round, and most acute in inter-harvest months

Outside of harvest periods, affordability of food and access to markets is particularly important. We have found that hunger has more to do with poverty than the sheer absence of food. For instance, most food sold in Nakivale is imported from further afield (e.g., Mbarara),¹⁴ but food frequently goes to waste when it cannot be sold.¹⁵

Affordability also impacts the diversity of foods that households can eat, and monotonous diets are known to cause micronutrient deficiencies.¹⁷ When families can afford food, they often stick to cheap single-grain diets.¹⁸ Some organisations suggest this is exacerbated by a lack of knowledge about the value of diverse diets,¹⁹ as well as cultural taboos restricting women and children from certain foods.²⁰

In response to hunger in the settlements, several grassroots organisations that we spoke to run programs distributing faster-growing crops, and help individuals join savings groups to tide them through the lean period.²¹

Hunger is stressful, and coping mechanisms are incredibly bleak

Account three — Kenya

In Kenya we heard from a researcher that comparable cuts to food assistance in a large settlement led to households selling the steel roofs of their houses, sending children to work for others, and resorting to transactional sex. In roughly 150 qualitative interviews conducted before assistance was cut, suicide came up once. In around 70 interviews afterwards, it came up almost every time.

New arrivals are a particularly vulnerable group

Settlement populations range from refugees who have lived there for many years to households who arrived only weeks ago, and how long a family has been in the settlement strongly shapes how well it can cope.

We learnt that new arrivals face significant disadvantages. They arrive having “left everything… [and] start from zero”.²⁴ They often arrive with little or no support network,²⁵ minimal English (Uganda’s official language, and often needed to access services and formal work), and no source of income.²⁶ They may also find themselves allocated land to reside on that is far from economic activity.²⁷

This is made worse by the fact that the support that new arrivals receive has dropped dramatically in recent years. New arrivals used to receive a full ration for 6 months (already shortened from 1 year) and a plot of land that was at least 30 x 30m (historically as large as 1-2 acres).²⁹ They are now folded into the prioritisation scheme, and although classed as ‘most vulnerable’ on entry, most get removed from this category swiftly (within 3 months,³⁰ though we have heard this is sometimes shorter in practice).³¹ The land they receive is also “now more like 10 x 15m”, or in some cases “30 x 30 feet”.³²

‘First-harvest-gap’

In line with its self-reliance policy, Uganda gives refugees the right to work and allocates households land within settlements.³³ Agriculture is an important part of this model: households are expected, over time, to produce some of their own food and rely less on humanitarian assistance.

The problem is that the timeline for food assistance does not necessarily match the timeline for becoming self-reliant. New arrivals currently receive full food assistance for 3 months before being moved into the wider prioritisation system. Yet even under favourable conditions, a household that receives land and plants immediately is unlikely to harvest staple crops before around the 4-month mark.³⁴ We call the period between the reduction in food assistance and a household’s first viable harvest the “first-harvest gap.” During this period, a household may have lost much of its food support without yet having had a realistic opportunity to replace it through its own production.

In practice, the gap may extend well beyond the theoretical 1-month window. Our interviews suggest that, in some cases, full assistance may end before the 3-month mark,³⁵ while a first harvest can take considerably longer than 4 months. A family arriving during the dry season may have to wait months before it can plant at all.³⁶ Families may also receive land late, lack suitable seeds or farming knowledge, or be allocated land whose productivity has declined after years of intensive cultivation.³⁷ Unpredictable rainfall and crop pests are additional ways a harvest can be delayed, or worse — wiped out.³⁸ Together, these factors can widen the “first-harvest gap” considerably.

There is also a more fundamental constraint: even reaching the first harvest does not necessarily mean reaching food self-sufficiency. We heard from farming experts that even the longstanding allocation of a 30 x 30m plot would not be sufficient for a family to meet all of its food needs.³⁹ The plot must accommodate a house and toilet, on top of crops, and is shared by a family of 6, on average.⁴⁰ One NGO worker told us that this amount of land might produce enough green vegetables, but not enough calories or protein to sustain a household.⁴¹

Our concern, therefore, is not simply that crops take longer than 3 months to grow. It is that food assistance can fall according to a fixed timetable, while households’ ability to replace that assistance depends on a much less predictable process. We want to understand how large this gap is in practice, which households are most exposed to it, and where support could most effectively prevent families from falling through it.

Food availability, affordability, and waste

We heard from one interview that food in Nakivale brought by vendors often goes to waste because people are unable to afford it.⁴² This struck us as particularly noteworthy because it suggests the problem lies less in food availability than in affordability. Many programmes focus on increasing food production, when limited purchasing power may be the more binding constraint. It also raises a hypothesis worth testing: if vendors lost less stock to spoilage, they might be able to sell perishable food more cheaply, or sell surplus at a discount rather than discard it.

Furthermore, many grassroots programmes produce food, but struggle with spoilage. Some crops are farmed intensively, but a bumper harvest might go to waste without storage or processing options.⁴³ Cold storage is generally difficult in Ugandan refugee camps,⁴⁴ but drying foods to prolong shelf-life has had some success.⁴⁵ Better storage and processing could let families stretch their harvests further into the lean season, shortening the period in which they go without enough food.

We would like to understand how accessible storage and processing facilities are to households and businesses in the settlements, and whether better access could help reduce spoilage and extend food availability into lean periods.