Delivering Resilient Enterprises and Market Systems (DREAMS) – Impact Evaluations

Decision Maker’s Challenge

For decades, the global response to refugee crises has focused on meeting urgent needs—providing food, water, and shelter. Yet displacement is increasingly long-term, leaving many refugees in camps for years with limited opportunities to earn income or rebuild their lives. As displacement becomes more protracted, many refugees spend formative stages of their lives in camps, raising concerns about the long-term and intergenerational consequences of limited economic opportunity.

At the same time, humanitarian funding is declining globally, placing additional strain on an already overstretched system. As resources dwindle, it has become increasingly clear that traditional humanitarian responses cannot sustainably support displaced populations indefinitely. This raises a critical question for humanitarian leaders: how can programs move beyond short-term aid to support sustainable livelihoods for displaced populations?

Ethiopia and Uganda host two of the three largest refugee populations in Africa, with approximately 1.1 million refugees in Ethiopia and 1.8 million in Uganda. As displacement grows while humanitarian funding declines, governments and donors need evidence on approaches that enable refugees and host communities to generate income and strengthen local economies.

The Delivering Resilient Enterprises and Market Systems (DREAMS) consortium, led by Village Enterprise, Mercy Corps, and IDinsight, tests whether combining poverty graduation with market systems development can help refugees and host communities build sustainable livelihoods in areas with fragile or thin markets. The program provides participants with business and financial training, mentorship, and market opportunities designed to support economic self-sufficiency. IDinsight is evaluating DREAMS in two refugee settings, West Nile in Uganda and Dollo Ado in Ethiopia.

Impact Opportunity

DREAMS tests a model that pairs household-level business support with efforts to strengthen local market systems, helping participants start businesses while also improving access to suppliers, buyers, and productive value chains. 

If successful, this approach could transform how refugee livelihood programs are designed and scaled, offering a pathway for displaced households to sustain themselves and contribute economically to the communities where they live.

Evaluating the DREAMS Program

We are conducting two large-scale randomized controlled trials (RCTs) in refugee settlements in West Nile (Uganda) and Dollo Ado (Ethiopia) to rigorously evaluate the DREAMS program. Each study uses a mixed-methods design, combining a household-level RCT with qualitative research to understand both the program’s impact and the mechanisms driving those outcomes. 

The evaluation focuses on three key questions:

1. What is the impact of DREAMS on the livelihoods, social cohesion, and perceived well-being of vulnerable refugee and host community households in West Nile and Dollo Ado refugee settlements?
2. What is the cost-effectiveness of DREAMS compared to the status quo?
3. [In Ethiopia only] What is the relative impact of DREAMS with and without direct market linkages, including input subsidies and value chain-specific support?

The evaluation includes two endlines: the first measures short-to medium-term impacts 1-1.5 years after implementation, and the second measures long-term outcomes 2-2.5 years after implementation.

Key Results

Program participation

High program participation: Almost all treatment participants in both contexts joined Business Savings Groups and formed Business Groups, the main elements of the PG component of DREAMS. In Uganda, 99% of participants joined Business Savings Groups, with 49% participating in a business in a promoted value chain, most commonly poultry and sesame. In Ethiopia, participation rates were similarly high (98% PG, 99% FD), with most Business Groups engaging in the shoat (sheep and goat) fattening value chain (90% and 94% in the PG and FD arms, respectively).

Economic outcomes

Higher consumption: Treatment households reported higher average monthly household consumption than control in both contexts, demonstrating a positive and statistically significant impact on economic welfare. In Uganda, consumption was higher by 17%, with gains observed across food and education expenditures. In Ethiopia, consumption was higher by 10% and 9% in the PG and FD arms, respectively, with no statistically significant differences between the two treatment groups.

Greater asset ownership: Treatment households reported higher total asset value than control in both contexts. In Uganda, asset values were higher by 21% and exceeded the value of program transfers, indicating that households were able to leverage business investments into asset growth. In Ethiopia, asset values were higher by 25% and 24% in the PG and FD arms, respectively, with no statistically significant differences between the two treatment groups.

Higher household income: Treatment households reported higher total monthly income than control in both contexts. In Uganda, income was higher by 24%, with larger gains observed among host community households. In Ethiopia, income was higher by 14% and 17% in the PG and FD arms, respectively, driven primarily by livestock-related activities.

Higher savings: Treatment households reported higher total household savings than control in both contexts. In Uganda, savings were higher by 108%, effectively doubling the savings buffer available to participating households, with gains driven in part by business savings. In Ethiopia, savings were higher by 92% and 91% in the PG and FD arms, respectively, similarly indicating a substantial strengthening of household financial buffers.

Resilience, financial inclusion, and well-being

Greater financial inclusion: The DREAMS program increased households’ use of financial services in both contexts. In Uganda, participation in community savings groups rose substantially (77% of treatment households compared to 44% of control households). In Ethiopia, mobile money usage increased from 68% among control households to 73% among treatment households, though uptake of formal banking remained limited.

Higher overall well-being: Treatment households reported modest, but significantly higher well-being in both contexts. In Uganda, well-being increased by 0.6 points on a 10-point scale, reflecting improvements in perceived happiness, health, and financial satisfaction. In Ethiopia, well-being increased by 0.2 and 0.3 points in the PG and FD arms, respectively, with a small but statistically significant positive difference in the treatment effect for FD households.

Improved food security: Treatment households reported modest but meaningful improvements in food security in both contexts. In Uganda, treatment households were 5 percentage points less likely to experience moderate or severe food insecurity than control households (69% of whom fell into this category). In Ethiopia, treatment households were 6 and 7 percentage points less likely in the PG and FD arms, respectively, than control (32% of whom fell into this category). While no significant improvements in children’s food security were observed in Uganda, in Ethiopia, children in PG and FD households were 6 and 9 percentage points less likely to skip meals than control, respectively (33% of control children skipped meals).

Greater economic empowerment for women: Female treatment respondents reported modestly higher levels of economic empowerment in both contexts. In Uganda, the women’s empowerment score was higher by +0.06 points on a scale from 0 to 1 relative to female control respondents (+0.54 standard deviations). In Ethiopia, women’s empowerment scores were higher by +0.03 and +0.04 points in the PG and FD arms, respectively, on a scale from 0 to 1 relative to a control mean of 0.26. DREAMS did not produce meaningful changes in social norms in either context, suggesting that harmful beliefs about women’s economic agency and mobility persist despite improvements in participation and decision-making.

Standardized effect sizes for our main indicators are shown below:

Cost-effectiveness: If effects are sustained for five years, the DREAMS program in both Uganda and Ethiopia will yield two to three times the value of its costs through increased household consumption and asset accumulation. In Ethiopia, the poverty graduation arm returned a marginally higher value per dollar invested than the full DREAMS arm, reflecting the modest incremental cost of direct market linkages relative to the similar welfare gains observed across both arms. Assuming modest positive spillovers, primarily from market systems development activities, would further increase BCR. The program’s cost-effectiveness compares favorably with peer programs in sub-Saharan Africa, such as the Village Enterprise DIB program (1.84x) and AVSI Rwamwanja (0.76x). Endline 2 results will provide further evidence on the sustainability of these impacts and the extent of spillover effects.

Read the results from Uganda and Ethiopia in our endline reports on Delivering Resilient Enterprises and Market Systems: