Every day counts: evaluating the benefits of early action during a crisis
Author: Lydia Poole
Photo: Talukdar David / Shutterstock
Typically, help in response to a humanitarian crisis comes after the fact and is requested through formal appeals. This type of response can be delayed, influenced by political factors, and often inadequate. UNOCHA and partners piloted a large-scale collective anticipatory humanitarian response to flooding in Bangladesh in 2020. In this insight, Lydia Poole, Associate Director for Evidence, shares her thoughts on what the evaluation of the intervention could mean for a system-level shift towards anticipatory response.
In 2020, the Centre for Disaster Protection partnered with the University of Oxford and the United Nations Office for the Coordination of Humanitarian Affairs (UN OCHA) on the first large-scale evaluation on the impact of anticipatory humanitarian cash transfers received in advance of a sudden climate shock. Specifically, in the context of extreme flooding along the Jamuna River in northern Bangladesh – the second highest and most protracted floods experienced in Bangladesh in more than 30 years.
Both the type of anticipatory humanitarian response – the fastest cash allocation of the UNOCHA Central Emergency Response Fund (CERF) in its history – and the evaluation itself were novel and challenging. The evaluation has recently been awarded a prestigious ADB-IEA Innovative Policy Research Award for showcasing an innovative policy tool that has high relevance to the operations of the Asian Development Bank, such as its contingent disaster financing mechanism recently piloted in the Philippines and Indonesia. The study was presented at the Asian Development Bank Annual Meetings.
Lessons from the Bangladesh anticipatory cash transfer study
The evaluation demonstrated a notable impact on food security, with households 52% less likely to go a day without eating if they received than households that did not. These effects on food consumption were present even three months after the cash transfer: Children in households that were sent the cash were 3.8% more likely to have consumed three meals in the day prior to the survey. It also found evidence that timely cash support went beyond providing temporary short-term relief during the peak of the flood; it also supported households to take action to reduce the wider impacts: households who received cash were 12.5% more likely to evacuate and 24% more likely to evacuate their livestock. The evaluation provides operationally relevant insights – notably, that the timing of response matters, with a small difference in adult food consumption found among those receiving the transfer just one day earlier. In order to respond early and fast however, requires upstream investments in targeting and preparedness.
Impact evaluations of anticipatory humanitarian actions present both practical and ethical challenges. Most notably because it is not certain if, when and, in some cases, exactly where an anticipatory response will be triggered. This ambiguity is combined with short lead times between triggers being activated and response. It is much harder to undertake the usual preparatory work, including identifying target beneficiaries and taking the necessary steps to set up a research study, such as conducting baseline surveys. The absence of baseline data requires large ex-post surveys. The impacts of one-off interventions are immediate, which in turn necessitates speed in data collection. This may mean using data collected via phone surveys. However, perhaps most significant is the challenge of identifying a valid control group in a context where randomly assigning households to treatment or control groups is seen to conflict with humanitarian principles. The Bangladesh anticipatory cash transfer study navigated these challenges using a quasi-experimental design to provide a counter-factual and a large post-intervention survey of nearly 9,000 flood-affected households (including both people who received the cash transfer and those who did not), to see how receiving this cash enabled them to better prepare for the crisis and protect their families and livelihoods.
The evaluation makes an important contribution to the evidence base supporting arguments for the welfare benefits of early action. It demonstrates that evaluating the impacts of anticipatory action during a crisis is challenging but possible.
Beyond a momentary boost to a lasting impact
The study points toward a next generation of research and operational challenges that anticipatory humanitarian action faces as its supporters seek to shift from pilots and build the case for anticipatory action to go to scale, and increase implementation in more challenging contexts.
For example, while it was possible to demonstrate the benefits of the transfer even months after the programme, the size of the impacts was, ultimately, small. Cash transfers provided two weeks’ support at the peak of the floods. Yet, the resulting crisis lasted far longer, and the cash transfers were not designed to dovetail with a post-shock response that could have better protected and built on the benefits of the early response.
Integrating effective anticipatory action into routine humanitarian programming – including risk and needs analysis, targeting, programme design and financing – will require major investments, mindset shifts, experimentation, and learning.
Experimentation with new financing instruments and models to enable more predictable financing for anticipatory humanitarian action is underway, with, for example, the IFRC adding a layer of insurance to its Disaster Response Emergency Fund (DREF) and OCHA and ARC exploring anticipatory insurance. Yet the more prosaic challenges of funding essential preparedness – including developing registries and early warning systems, agreeing plans and pre-positioning supplies – alongside recovery and longer-term resilience building, are substantial and difficult for advocates of anticipatory humanitarian action to influence alone.
Moving from pilots to scale and addressing implementation in challenging contexts
Adapting the humanitarian programming toolbox to enable anticipatory action in fragile and conflict-affected settings will require substantial learning and investment. Anticipatory action in Bangladesh was already a well-established site of innovation prior to the climate shock-specific cash transfers, and Bangladesh is a politically stable environment. Risk modelling data exists, triggers and programming responses had been tested, and a large-scale response was feasible.
It is far less clear what anticipatory actions make sense in complex and protracted crises with multiple overlapping shocks, frequent population movements and limited reliable forecasting data. For example, in South Sudan, in exploring the potential for an anticipatory action pilot, CERF and partners concluded that since predictive performance of global forecasts was poor in South Sudan and national forecasts were unavailable, a standard anticipatory action approach would not be technically feasible. Recent research from the Supporting Pastoralism and Agriculture in Recurrent and Protracted Crises (SPARC) programme in Somalia, which sought to identify relevant anticipatory actions and timelines, found instead just how difficult achieving this seductively simple ‘stitch in time’ logic and the clarity of agreed triggers, proves to be in the context of ‘wicked crises’, where impacts of multiple shocks vary widely, and the timing and content of anticipatory actions must be highly contextualised to be effective. Ironically, it was the 2011 famine in Somalia, in which more than a quarter of a million people died, that provided a major impetus for humanitarian actors to rethink funding and response models, including anticipatory humanitarian action.
The next generation of challenges for meaningful and effective anticipatory humanitarian action is immense, but so are the potential rewards. Confronting these challenges brings transformative potential to how the most at-risk communities, governments – and the international humanitarian system – can better prepare for and respond to shocks. Investing in well-targeted, high-quality and publicly available evidence significantly accelerate this shift.