Flying blind - insight gaps for evidence-based prioritisation

The recent funding cuts have severely weakened the system’s analytical capacity. This has resulted in a ‘looming evidence crisis’.^[1] This year has seen a massive loss of dedicated analytical staff and a diminished operational footprint of key data-gathering actors, and it has revealed the systemic vulnerability of reliance on single donors (primarily the US) for essential data ‘enablers’.

Capacities to understand and monitor people in need and evolving crises are under serious threat. When USAID cut its funding to the famine early warning system FEWSNET at the start of 2025, many saw it as presaging a worsening ‘data drought’ (TNH, 2025). The interconnectedness of the information system means that the loss of critical sources has a cascading effect – and this has prompted concerns that humanitarian response plans in some countries are running on outdated information.

Humanitarian planning and response rely on a highly interconnected information ecosystem (Simons, 2025), which has suffered funding precarity. This ecosystem includes the Multi-Sectoral Needs Analysis (MSNAs) conducted by REACH, an independent organisation. IOM’s Displacement Tracking Matrix (DTM) is another ‘critical enabler’ – its primary data collection underpins analysis in up to 95% of recent HNRPs and it feeds into the Index for Risk Management (INFORM) used in several donors’ decision-making. IOM staff reported that around half of its operations were seriously impacted by the USAID aid cuts. While some were temporarily secured, all 80 of the active DTM operations are under serious stress and three quarters face possible closure in 2026 without additional funding.

The reduction in monitoring systems, including lower frequency of surveys already seen in 2025, raises the risk of the system failing to detect existing needs and emerging crises. This includes famine risk and disease outbreaks – hampering the prospects of early action. The implications extend to the political and funding realms too, as the robustness of humanitarian analysis comes under increasing scrutiny by overstretched donors, and by governments seeking to present counter-narratives to deny or downplay crises.

Donors have backfilled some gaps, but the outlook is uncertain. Recognising the growing problem, early in 2025 the Swedish International Development Agency (SIDA) convened a group of donors to focus on support to these critical enablers. Individual donors reported stepping in with bridge funding over the year to keep services running. Proposals have been mooted to fund data enablers through global funding models but, given that the overall level of funding is shrinking, the administrative costs of this have been questioned.

Putting communities at the centre emerged as a priority at the policy level but moving beyond performative consultation is a matter of intense debate. As the aid cuts took hold, there was both a return to headline commitments to consult populations affected by crisis and, conversely, a retreat from putting this commitment into practice. As one source put it, ‘Humanitarian Country Teams are having to make cut-throat decisions and yet wanting to appear as people-centred as possible’.

In line with the Emergency Relief Coordinator’s Reset call to ‘put people facing crises first’ (OCHA, 2025e), OCHA notes that ‘while there has been growing engagement of communities in the HPC – between HPC 2023 and HPC 2025, the percentage of countries where people were consulted during response planning increased from 68% to 85% – this has not gone far enough’ (OCHA, 2025c). However, the proposals for the 2026 Humanitarian Programme Cycle (HPC) process set this within narrow parameters – calling for use of agencies’ existing community engagement to identify preferences and wherever possible selecting available response modalities that align best with these. This has been met with mixed responses – some recognise the limited room for manoeuvre, and many others are concerned that it amounts to ‘kicking the ball down the road’^[2] to agencies who are decreasingly resourced and inclined to prioritise it. One source noted ‘there is this misunderstanding that community consultation is something that the system can’t afford at this stage’, rather than something that humanitarians can ill-afford to neglect.

Experiences have differed between countries – an agency representative expressed concern that, in South Sudan, there was ‘no room’ in the HNRP for the findings of community consultations. In both Somalia and Colombia, meanwhile, country teams have been working to strengthen community engagement and use decentralised approaches to coordination under the Flagship Initiative, which may provide models for new approaches to community engagement even under resource constrained conditions. In Somalia, for example, the Humanitarian Country Team has engaged Ground Truth Solutions to review its secondary community consultation data with a view to using it better. And in Colombia, harmonised and agreed-upon tools for community consultation have been developed, and the use of artificial intelligence to synthesise data and information is being accelerated (SIDA et al, 2025).