Do Livelihood Programs Really Help Displaced People? A sweeping review finds thin evidence

For nearly two decades, humanitarian organizations have pinned considerable hope on a package of interventions known as the Graduation Approach, a sequenced livelihoods program designed to lift the world’s most vulnerable households out of extreme poverty. Originally pioneered by BRAC in Bangladesh, the model bundles together asset transfers, training, coaching, and access to savings or credit, with the idea that each element builds on the last until a household can stand on its own. In recent years, the approach has migrated from development settings into the far messier world of displacement, where refugees and internally displaced people face legal restrictions on work, precarious housing, and chronic dependence on aid. A new systematic review now asks a deceptively simple question: does the model actually work for displaced populations? The answer, according to researchers publishing in PLOS Sustainability and Transformation, is that the evidence base is far thinner than the enthusiasm surrounding the programs would suggest.

The review, led by Lauren N. Yan and colleagues, screened 575 records drawn from eight academic databases and the websites of eleven implementing agencies, covering literature published between 2002 and 2024. The team searched EconLit, PubMed, Web of Science, the Cochrane Library, and several specialized databases in the fields of psychology, global health, and family studies, with the final search updated in October 2024. From that large pool, only five studies met the inclusion criteria, together covering 28,873 displaced adult participants. That ratio, five usable studies out of nearly six hundred records, is itself a striking finding. For an intervention model that has been widely adopted by humanitarian actors across the globe, the rigorous evidence trail in displacement contexts is remarkably sparse.

The five included studies were geographically concentrated in four countries: Uganda, Mozambique, Colombia, and Costa Rica. They spanned a methodological range from randomized and non-randomized trials to qualitative studies and program reports, and the review team assessed their quality using the Mixed Methods Appraisal Tool, a validated instrument for systematic mixed-studies reviews. The research was registered in advance on PROSPERO, the international prospective register of systematic reviews, under protocol CRD42023387899 in May 2023, a step that helps guard against post hoc changes in how the evidence was evaluated. Even so, the authors report that the studies provided insufficient detail to fully assess their quality or risk of bias, a limitation that shadows every conclusion drawn from them.

What did the five studies show? On the most basic measures, the picture was cautiously positive. Participants in Graduation-style programs generally experienced improvements in food security and living conditions, suggesting that the core mechanics of the model, providing productive assets and support to displaced households, can produce tangible short-term gains even in difficult settings. But the review found that these changes were rarely quantified. Programs reported that things got better without specifying by how much, for whom, or relative to what baseline. In evidence terms, that is a significant weakness: without numbers, it is impossible to distinguish a meaningful intervention effect from the natural drift of circumstances, seasonal variation, or the general tendency of participants and implementers to report success.

Employment outcomes, arguably the heart of any livelihoods program, were reported only rarely. For displaced people, who often face legal barriers to formal work and deep discrimination in labor markets, whether a program leads to actual jobs or self-employment income is the question that matters most. The review found that social wellbeing, mental health, and poverty outcomes were heterogeneous or unclear across studies, meaning the results pointed in different directions or could not be confidently interpreted. Perhaps most strikingly, the review found that dependency on humanitarian aid, a central concern in displacement policy and a key rationale for livelihoods programming in the first place, was not directly assessed by any of the included studies. The very outcome that motivates much of the investment in these programs has essentially never been measured.

The heterogeneity ran so deep that the authors could not pool results statistically. Substantial differences in how outcomes were defined and measured precluded any quantitative comparison or meta-analysis, the technique that allows reviewers to combine effect sizes across studies and arrive at a precise overall estimate. Instead, the team conducted a narrative synthesis, describing patterns in words rather than numbers. This is an honest methodological choice, but it also means the field lacks the kind of aggregated evidence that has made Graduation Approach evaluations in stable, non-displaced settings, such as the multi-country randomized trials published in Science in 2015, so influential. Displacement contexts, it turns out, have been evaluated with a fraction of that rigor.

On implementation, the review offered more textured insights, though they come with caveats. Barriers and facilitators were found to be complex and context-specific, defying any one-size-fits-all prescription. Still, one theme emerged as central: comprehensive community-level support appeared to be key to program success. Programs that engaged the communities hosting displaced participants, rather than treating beneficiaries as isolated households, seemed better positioned to succeed. Flexibility in implementation also emerged as critical for feasibility, allowing programs to adapt to shifting conditions such as changing regulations, market disruptions, or new waves of displacement. In other words, the Graduation Approach’s famous sequencing may matter less in displacement settings than the ability to improvise around it.

Sustainability, the question of whether gains persist after a program ends, proved especially sensitive to structural factors beyond any organization’s direct control. Land tenure was a recurring concern: displaced participants who lack secure rights to land cannot reliably farm, keep livestock, or build the fixed assets that many livelihoods strategies assume. Market inter-reliance was another fragility, since small enterprises in displacement settings often depend on the same saturated local markets as their neighbors, meaning one household’s success can come at another’s expense or collapse when local demand shifts. These findings underline a sobering point about humanitarian livelihoods work: the binding constraints are often legal and economic structures, not the motivation or capability of the participants themselves.

The authors are careful not to claim that Graduation programs fail in displacement contexts. Rather, their conclusion is that despite the positive conclusions stated by the reviewed studies, there is limited implementation insight and little robust evidence of effectiveness. That distinction matters. The gap between what implementers report and what the evidence can support is precisely the kind of finding that should prompt reflection among donors and agencies. With millions of people in prolonged displacement worldwide and humanitarian budgets under strain, the case for investing in livelihoods programming is strong on principle, but principle is not evidence, and the review suggests the evidence has not kept pace with the programming.

The path forward, the researchers argue, requires better studies, not necessarily different programs. Future research should address counterfactuals, meaning proper comparison groups that reveal what would have happened without the intervention, along with long-term outcomes that track participants for years rather than months, and implementation questions that explain how and why programs succeed or fail in specific contexts. Standardized measurement of outcomes such as aid dependency, employment, and mental health would allow future reviews to pool results and draw firmer conclusions. Until then, the Graduation Approach remains a promising but under-evidenced tool for displaced populations, a model whose global momentum has outrun the data needed to justify it. For a field that prides itself on evidence-based humanitarianism, that is a gap worth closing.

Subject of Research: Effectiveness and implementation of Graduation Approach livelihood programs for displaced populations

Article Title: Effectiveness and implementation of “Graduation Approach” livelihood programs with displaced populations: A systematic review

Article References: Yan, L. N., Wang, Y. E., Mubangizi, F., Parks, C., Whittemore, C., & Zelelke, N. (2026). Effectiveness and implementation of “Graduation Approach” livelihood programs with displaced populations: A systematic review. PLOS Sustainability and Transformation, 5(6), e0000252. https://doi.org/10.1371/journal.pstr.0000252

Image Credits: AI Generated

DOI: 10.1371/journal.pstr.0000252

Keywords: Graduation Approach, displaced populations, systematic review, livelihoods, humanitarian aid, food security, refugees, poverty alleviation, implementation science, PLOS Sustainability and Transformation, evidence gaps, sustainability