Acting early — the causal impacts of WFP’s anticipatory action programmes in Nepal and Bangladesh
By and
As climate-related shocks grow in frequency and intensity, humanitarian actors face increasing pressure to improve both the timing and effectiveness of their responses.
In 2023 alone, extreme weather events were the main driver of food insecurity for an estimated 72 million people across 18 countries.
In this context, anticipatory action — delivering assistance before a disaster strikes — is emerging as a critical tool. But how effective is it? What trade-offs are involved? And what outcomes can be expected compared to traditional post-shock responses?
The World Food Programme (WFP), in collaboration with the World Bank Development Impact group, has been testing these questions through a series of impact evaluations, focusing on WFP flood anticipatory action programmes in Nepal and Bangladesh. In this blog, we summarize key findings and what they imply for shock-responsive programming.
Defining anticipatory action
Anticipatory action refers to pre-arranged interventions — such as early warnings, cash transfers, or other essential support — triggered by forecast-based systems before a crisis unfolds. To function effectively, anticipatory action relies on:
- Agreed-upon trigger thresholds (for example, rainfall or river levels)
- Prepositioned funding and partners
- Anticipatory action plans that can be rapidly implemented ahead of a climate hazard
The central hypothesis is that timing matters: reaching households ahead of a climate shock could mitigate harm and reduce reliance on negative coping mechanisms.
WFP designed impact evaluations that were ready to be deployed whenever a severe flood event would hit either Nepal or Bangladesh. Action triggers were met in Nepal in 2022 and in Bangladesh in 2024.
Nepal
In Nepal, WFP implemented a randomized controlled trial during the 2022 monsoon season in the Karnali River basin. Once flood triggers were reached in early October, cash assistance (USD 117 per household) was distributed to around 12,500 households.
Two groups were compared
- Anticipatory action group: Received cash transfers immediately after triggers were activated
- Post-shock group: Received the same cash 3-4 weeks later, after peak flooding and initial damage assessment
Impacts in the short term
- Food security: Anticipatory action recipients showed significantly higher food consumption, especially of animal protein
- Coping strategies: Households that received anticipatory action support were less likely to borrow food, skip meals, or reduce portion sizes
- Psychosocial wellbeing: Life satisfaction and mental health scores improved relative to the post-shock group
While the post-shock group eventually received the same transfer, they did not catch up in any of the measured outcomes — suggesting that anticipatory action generated net welfare gains, not just timing shifts.
Bangladesh (Report forthcoming)
After anticipatory action triggers were activated in five districts in the Jamuna basin in July 2024, WFP Bangladesh reached almost 90,000 households through mobile payments at least one day ahead of the flood. A second impact evaluation took place, where villages were randomly assigned to one of three groups:
- Anticipatory action group: Received USD 43 one or two days before the flood
- Rapid post-shock group: Received the same amount about 3 weeks after peak flooding
- Standard post-shock group: Received the transfer 8 weeks after peak flooding
Findings
- The anticipatory action group reported better food security, fewer coping strategies, and improved mental health compared to both post-shock groups.
- These differences were most pronounced in the early rounds of data collection, gradually converging over time but still visible up to three months later.
Interpreting the evidence
The studies suggest that timing alone — holding the amount and type of assistance constant — can significantly affect short- and medium-term outcomes. In both settings, anticipatory action led to:
- Higher food consumption
- Reduced use of negative coping strategies
- Improved subjective wellbeing
The results also speak to broader questions in humanitarian design: the trade-off between timeliness and targeting accuracy, the cost-effectiveness of anticipatory disbursements, and the operational feasibility of anticipatory action systems.
Implications for future programming and research
As of 2024, WFP’s anticipatory action programming has expanded to over 44 countries, reaching 6.2 million people. Upcoming impact evaluations will cover other hazards, such as droughts and cyclones, with a focus on:
- Cost-benefit analysis of anticipatory action versus post-shock models
- Targeting refinement to ensure the inclusion of high-risk populations in anticipatory action programming
- Analyzing different anticipatory action assistance packages (for example, early warnings plus agricultural inputs)
Going forward, the key question is not whether anticipatory action works — the evidence increasingly suggests it does — but how to optimize its design, scale it sustainably, and adapt it across different shocks and contexts.
See also the short Brief, summarizing the findings from both countries linked here
