What are the current problems faced by humanitarian actors looking to effectively communicate their impact? How can we improve the way that evidence is generated in the humanitarian sector?
Recent aid cuts have shocked the world, both because of their speed and sweeping global impacts. USAID alone is estimated to have saved approximately 3.3 million lives per year, highlighting some of the severe negative consequences we can expect to see.
Aid cuts have been largely underpinned by two misconceptions of aid and its impact. First, that aid is ineffective. Mainstream media has played a powerful role in shaping negative narratives about aid, often amplifying stories of failure - such as high-profile cases of mismanagement - whilst underreporting the long-term successes of locally driven solutions. This has helped to form negative perceptions about aid - polling shows that people believe it is mostly wasted on corruption and overestimate the total amount of aid funding.
Second, that aid is a distraction from domestic priorities that are becoming more urgent and complex to address, some of which have been longstanding such as building military and defense capabilities. In Britain, the 2024 Survey of UK Public Opinion on Foreign Policy and Global Affairs shows that Britons are highly sceptical of the benefits of the UK’s international development and aid programmes– 56% don’t believe it helps tackle irregular migration and 45% don’t believe it makes the UK safer. In turn, only 11% of Britons want to see an immediate return to spending 0.7% GNI on aid.
Those negative perceptions of aid call for an urgent rethink of how we evidence and communicate effectiveness within the aid sector. In this article, drawing on our experience within the humanitarian sector, we focus on key learnings and approaches in evidencing impact. We also outline a framework for evidence and learning that will continue to inform our research and practice within the sector.
Three problems in evidencing impact
A persistent and unhelpful overreliance on quantitative data - Humanitarian organisations have historically faced pressure to demonstrate the effectiveness of their programmes and practices in addressing urgent needs. As a result, there has been a persistent focus on using statistically measurable outcomes, often at the expense of understanding the underlying reasons why certain practices have succeeded or failed. With qualitative and mixed methods analysis relegated to a more peripheral space in evidence generation, a more holistic understanding of how and why change happens when evaluating outcomes is often missing. This is particularly relevant to the growing interest in evidencing effectiveness and value for money (VfM) which tend again to focus on numbers (such as number of lives saved) rather than the less quantifiable but equally important aspects of a humanitarian response (contributions to resilience, dignity and so on).
Limited engagement of local actors – Evidencing impact is still very much Northern-led despite calls for localisation of research design and implementation. There is still a view that experts from the Global South are “emerging” or less rigorous than Western or northern experts. This imbalance creates a contradiction in localisation: while local organisations are increasingly expected to deliver humanitarian programmes, they are often excluded from shaping how their own impact is assessed. As a result, success continues to be measured through externally imposed frameworks that may not reflect what matters most to affected communities.
Ethical challenges and methodological limitations in humanitarian research - Research within the humanitarian sector often faces significant ethical and methodological challenges, driven by a combination of donor expectations and the limited infrastructure typical of humanitarian and conflict-affected contexts. As a result, core ethical principles—such as obtaining informed consent, safeguarding participants, and engaging meaningfully with local actors and experts—are frequently sidelined. Moreover, the donor-driven emphasis on producing quantifiable results has shaped how impact is assessed. Under pressure to meet reporting requirements, local actors may generate numerical data that is arbitrary or detached from the complex realities on the ground. This over-reliance on quantitative metrics has undermined methodological rigour, restricting opportunities to examine the historical, systemic, and contextual dimensions of change.
Moreover, evidence and learning are often considered secondary support or enabling functions to a humanitarian response rather than an anchor for stronger, more effective humanitarian engagement, and an opportunity for leveraging local experiences and needs with much more clarity. A 2015 paper reviewed impact evaluations of humanitarian assistance conducted since 2005 and found several significant shortcomings: of 39 papers reviewed, 23 failed to assess whether their comparison groups were valid; 29 did not evaluate the precision of their results and only five addressed ethical considerations. These gaps highlight the need for a shift towards more robust, contextually grounded approaches to evidence generation and learning.
How can we improve the way that evidence is generated in the humanitarian sector?
Adopting co-production and localisation of design and analysis - Actively engaging local actors in research and evaluation design, data collection and dissemination. This active co-production and localisation of the research process enhances its robustness and accuracy. When context-specific metrics of impacts are considered, it also improves accountability to populations at-risk of, and affected by, crisis, as they are able to shape the determinants of success rather than have them imposed solely from Western donors or INGOs. Our own experiences highlight the importance of involving local actors as equal partners in design and data collection—rather than treating them merely as sources of information.
Centring local voices and building research capacity – We need to move away from methodologies that prioritise Western expertise and instead promote approaches that centre local voices, experiences, languages, and leadership. This includes, for example, embedding a decolonising approach in the way we design our research projects, which may involve the greater use of local languages, the presentation of data in non-Western formats, and adequate time and resource spent addressing power dynamics that might exist between the research team and participants. Beyond improving the way we collect and communicate data; we need to work to strengthen the research capacity of local organisations.
Embedding innovative and multidisciplinary methodologies - With the growing momentum around multidisciplinary research, by integrating both social and hard sciences, as well as methodologies from fields like public health and climate science, we can deepen our understanding of the current and future landscape of the humanitarian sector. New collaborations are required to reimagine the future of humanitarian work, particularly in relation to locally-led action, while fresh approaches to measuring value for money and improving the efficiency of aid delivery need to be developed.