Efforts are underway to advance connectivity through the expansion of telecommunication services and automated warning messages. However, such improvements will be insufficient unless more accessible guidance is made available on how individuals and communities can prepare. The EW4All initiative’s global scope lends itself to high-level policy interventions, which are essential but cannot in themselves address the gaps in communicating potential impacts that emerge early in cascading, cross-border risks (for example, in relation to disruption to food supply or reduced food usability).
Structural gaps persist at the national level, where ministries, administrative units and other stakeholders fail to coordinate with one another in practice. Even when national monitoring is mandated, there is often no clear authority to issue food-related alerts. Risk communication typically falls under national disaster preparedness ministries, in partnership with service providers for network connectivity, and pieced together by international organizations drawing on fragmented data from national meteorological services, agricultural ministries, disaster preparedness units, district authorities and civil society. In the absence of localized, granular data and contextual analysis, global-level forecasts and models are used to fill gaps. However, existing tools and warning systems are optimized to detect extreme events, such as famine prevention, not the more subtle but critical signs of food insecurity, such as price volatility, disruptions of food access, reduced dietary diversity or reduced nutritious value of food, that precede severe outcomes.
The failure to monitor cascading risks compounds this issue. While initiatives such as the AMIS provide insight into global commodity markets, there is no end-to-end system that tracks how disruptions in trade, climate or health systems ripple through to affect food security at the household level. These disconnections were voiced in the Missing Voices interviews, where individuals in Bangladesh and Senegal shared experiences of repeated crop failures, livestock loss and disrupted food supplies with no warning or coordinated response. Key data are either missing or under-prioritized, impacting the timely issuance of an early-warning alert. The inability to embed cascading risk frameworks into multi-hazard early-warning system design can severely disadvantage timely, pre-emptive communication to mitigate risks before they escalate, as available early-warning systems fall short of analysing how impacts ripple between trade and food systems, or health and food systems.
Even when alerts are distributed well, hazard-specific warnings can be technical, which limits local comprehension and the ability of communities to understand what preparedness measures they must take. This hinders the effective use of meteorological data, even in single-hazard warnings, let alone where communities face multiple risks. For instance, flood-specific warnings are often too technical in Bangladesh, limiting local communities’ understanding of them and reducing their ability to take appropriate preparedness measures. The Bangladesh Missing Voices study found that fewer than 10 per cent of respondents consistently understood hazard-specific alerts. Similarly, in Senegal, lack of training and restricted access to technological tools affected the ability of local populations to understand warnings.
The gaps in warning dissemination and communication are often characterized as an absence of data but could more accurately be described as an absence of usable, trusted and timely data for decision-making at the community level. Many warning systems are more suited to diagnosing and alerting to shocks that have been experienced in the past, rather than anticipating potentially new future ones. Community-led warning and dissemination, where they exist, are good at localizing alerts, for example where local CSOs, such as Jokolanté in Senegal, are initiating weather alerts followed by voice calls in local languages to offer advice on preparedness actions. However, sustaining such initiatives over the long term is difficult due to the costs associated with this and the lack of recurrent funding.
Reconceptualizing people-centred early-warning systems
Currently, at the global level, progress in communication and dissemination of early warning focuses on improving digital connectivity and misses nuances such as potential disincentives to communicate risks at national and subnational levels – an example of this could be a perception among politicians that spending on visible relief efforts in the wake of a disaster may afford them greater acclaim than invisible and potentially redundant investment in improved warning systems. Furthermore, although dissemination channels for early warnings are diverse, they are hindered by community-level barriers to accessibility, according to focus group discussions in the case study countries. Some individuals may face barriers to accessibility to otherwise common dissemination sources due to race, gender, ethnicity or sexual orientation. Communities that are harder to reach may need additional dissemination efforts. The specific needs of those communities should be better accounted for in existing protocols, which require analysis of multidimensional vulnerabilities that can disrupt their ability to receive early-warning information before a disaster impact.
What emerges is a need for multi-hazard early-warning systems that are not just technically sound, but also inclusive, participatory and grounded in local realities. For example, the Missing Voices study in Bangladesh and Senegal highlights how unequal access to information deeply impacts food security, particularly for women-led households, older people and those with disabilities. These groups are frequently excluded from formal early-warning systems and must rely on informal sources of information that are not always reliable or timely. Prioritizing their access to information is crucial – not just in format or connectivity to mobile systems, but in timing, language, trust and delivery.
Practitioners of anticipatory action in international humanitarian organizations have long emphasized the importance of people-centred early-warning systems, and their insights are crucial for shaping inclusive risk communication. Therefore, our understanding of people-centred early-warning systems must move beyond relying on improved data and connectivity alone. Unlike approaches focusing on (single and multiple) hazards, embedding cascading risks methodology in early-warning systems with a focus on the impacts on people can account for the different ways individuals within communities may experience and respond to risks differently, such as to their food security. To be successful, this people-centred approach must ensure local voices are not just recipients of alerts, but active contributors to its development, interpretation and use will be key to building more inclusive, responsive, and ultimately effective warning and dissemination systems.
Pillar 4: Preparedness and response capabilities (plans, practices, resources)
Pillar 4 focuses on ensuring early warnings translate into timely, anticipatory actions that save lives and livelihoods. This requires the establishment of preparedness plans and operational protocols at local, subnational and national levels. In current early-warning systems, these protocols tend to be set at a national or district level with pre-agreed activities, intended to be nimble and adaptive, with reiterations and revisions made after each activation or every few years. They rely on accurate forecasts of impacts and agree on likely triggers for action to be able to inform and link to appropriate early or anticipatory action protocols or activities (at national, institutional or local levels). However, evidence from case studies in Bangladesh and Senegal reveals several systemic and contextual challenges that impact the effectiveness of preparedness and early action to food insecurity.
National policy frameworks increasingly consist of adaptation, resilience and immediate disaster response approaches. However, there remain frequent disconnects between these long-term strategies and the necessary immediate, early action and humanitarian relief. In practice, the emphasis tends to fall on short-term relief efforts, with limited integration of anticipatory approaches linked to food security and nutrition. Opportunities to align early action with broader food security strategies are missed, particularly in the context of transboundary and cascading risks that require systems analysis beyond local and direct impacts.
Effective early-warning systems, when properly designed, can interrupt cascading impacts before they affect vulnerable populations by making the most of the critical window between forecast and impact. Additionally, such systems can facilitate localized adaptation by reinforcing food production, supply chains and storage infrastructure, ultimately reducing vulnerability and bolstering community resilience.
Furthermore, most food security early-warning systems are structured around hazard detection, rather than impact forecasting. As a result, warnings may fail to reflect or communicate the nuance of food security risks at the household level. For instance, the inability to capture upstream disruptions in supply chains, informal food markets or livelihood assets means critical triggers for early action are missed. This is compounded by challenges in identifying diverse exposures and vulnerabilities in multi-hazard environments at the intersection of natural hazards and socio-economic stressors, such as food price spikes or reduced remittance flows, which can deepen food insecurity without triggering a formal alert.
Implementation plans and procedures for response to early warnings
At the national level, anticipatory action protocols establish early-action measures that are underpinned by finance that is ready to be allocated as soon as a certain – often quantitative – threshold is met or the ‘right point of intervention’ is reached, often easier to identify in sudden impacts like floods, than more slow and complex impacts.
When upstream risks are better monitored at the regional and national levels, communities could be given more lead time in preparing their communities’ resilience to food shocks – before a disaster.
Analysis of the complex food crises in Ethiopia by the Academic Alliance for Anticipatory Action found ongoing challenges in determining ‘the right moment’ for anticipatory action. These challenges stem from varying timescales for an effective early response, as what might be seen as an early intervention for one crisis, e.g. droughts or supply chain shocks, could be a delayed response to another, such as in coastal flooding, where communities must act very quickly. In Bangladesh and Senegal, some research participants indicated that if they received alerts on the likelihood of food security threats early enough, they could take some preventative measures, such as harvesting or securing food. When upstream risks are better monitored at the regional and national levels, communities could be given more lead time in preparing their communities’ resilience to food shocks – before a disaster.
Early action can be taken at multiple stages, depending on the scope and design of the early-warning system. For instance, systems focused on the ‘forecast-to-impact’ window may support a ‘block’ intervention to stop cascading effects before they reach a vulnerable community. Alternatively, these systems may enable a ‘domestic adaptation’ response to help communities absorb impacts by addressing underlying vulnerabilities. Other approaches, such as ‘adaptation at origin’, ‘adaptation along the transmission pathway’, or ‘system-wide adaptation’ intervene earlier in the disaster risk management cycle (see Figure 3), and while some of these fall beyond the remit of anticipatory action – from recovery to preparedness – these strategies aim to prevent or reduce cascading risks before they manifest at the community level.
While there is always a risk of false warnings, preparedness increasingly pays dividends in the long run. Corroborated data can help diversify and support decision-making. Already in some instances, even if the forecasted impact does not occur, earlier action to protect high-risk individuals is taken due to the lower cost and relative efficiency of acting early to save lives or livelihoods. Such actions, often referred to as low-regret triggers, should be reflective of the needs and vulnerabilities within communities. In some instances, advancements in real-time monitoring can bridge this knowledge gap with tools like WFP’s Vulnerability Analysis and Mapping (VAM) mechanism. However, where real-time monitoring lacks a connection to national or international food systems data, decision-makers may lack the necessary evidence to initiate a response.
All these systems are contingent on funding that is premised on the principle that acting early reduces the costs of response, saves lives and livelihoods. However, early warnings are often not converted into early action. Reasons for this can be diverse, including a lack of timely funding, governance or coordination gaps in implementing early-action protocols, disconnects between access to disseminated warnings or a lack of warnings themselves.
Furthermore, even when pre-arranged finance is available, disbursement of funding often depends on meeting specific quantitative thresholds. However, such thresholds are more effective in simpler contexts and fall short in complex, cascading risk contexts and in relation to existing vulnerabilities. Research by the Stockholm Environment Institute (SEI) suggests that in adaptation contexts, integrating ‘concern assessments’ alongside a technical risk assessment could enhance decision-making. These assessments can improve understanding of community perceptions on threats to food access, availability, usability and stability. This can help ‘systematic[ally] analys[e] …perceived impacts that communities associate with a hazard, its cause and consequences’. This allows space for distinctions between communities, those marginalized within them, and how they are affected differently by impacts.
For example, some types of food relief in anticipatory action may increase the food security of one group but undermine the food security of another due to the cascading and transboundary risk context, for example pre-positioning food aid in a location that distorts local markets and affects producers’ incomes. To avoid trade-offs and unintended impacts, early action often takes the form of cash and voucher assistance, allowing more autonomy for individuals to identify and meet their needs and reduce risks of maladaptation, exclusion from food relief and of individuals adopting coping mechanisms that are harmful to their long-term well-being.
Coordination and governance of stakeholders
Social and political considerations often delay the activation of early-action mechanisms, even when forecasts indicate a high probability of impact. When specific ministries or coordinating bodies fail to take responsibility for cascading risks, it can undermine accountability, trust in governments and discourage other stakeholders from stepping in. Such a lack of leadership also creates gaps between ministries and civil society, as well as between those issuing warnings and those who receive them. Communities, as a result, may lack the necessary information for their preparedness. In addition, trade monitoring mechanisms are largely global and regional in scale, often overlooking local realities. This disconnect hampers the trust in early-warning information. Whereas monitoring of acute food insecurity is usually better connected to communities, monitoring in real time is necessary to ensure timely decisions are made. In resource-constrained settings, food availability is often the focus of analysis on a population’s food security. Therefore, not all aspects of food security are analysed, creating knowledge gaps in access, usability and stability. Intricacies of building technically sound early-warning systems tend to steer focus away from understanding how well these systems work in practice.
Social and political considerations often delay the activation of early-action mechanisms, even when forecasts indicate a high probability of impact.
However, by centring on communities as end-users of warning information, existing information could be a bridge between global and national trade monitoring, regional and national hazard and exposure mapping, and across all levels up to national-level monitoring of vulnerabilities supported by international humanitarian organizations and civil society. While ample attention is paid to the generation of new, better-quality data, despite good efforts, more can be done on the coordination across existing pockets of data at the global, regional, national and community levels.
Market-based early-warning approaches, such as the Agricultural Market Information System (AMIS), offer insights into global price volatility and commodity trade flows. However, unlike acute food insecurity monitoring, they remain insufficiently connected to community-level risk monitoring. Therefore, communities may feel their food security is already affected before early warning systems record any impact, and this can increase vulnerability to adverse weather or other disruptions to food. In fragile or conflict-affected areas, this disconnect is particularly stark as data collection at the local level can be impeded by fighting and instability. Linking agricultural trade data with localized warning systems could help anticipate food security crises by diversifying data sources to plug any data gaps, where direct monitoring is constrained. But this requires expanded institutional mandates, flexible financing mechanisms and stronger integration between humanitarian, development and climate finance mechanisms. This is discussed in more detail in Chapter 4 of this paper.
Some countries have made progress in this area and the results illustrate the effectiveness of coordination and community-centred approaches. In Senegal, the national coordinating body on food security (Secrétariat Exécutif du Conseil National de Sécurité Alimentaire – SECNSA) is integrated into the regional Cadre Harmonisé framework, which successfully shares food security assessments with regional organizations, enhancing the effectiveness of early warnings. For example, during the 2017–18 Sahel drought, early warnings using satellite-based tools identified vulnerable regions in northern Senegal, facilitating targeted responses. However, the tool is currently better suited to slow-onset hazards, such as droughts, when compared to rapid-onset hazards as it is unable to capture vulnerabilities that interact with different risks in dynamic ways. Unlike Senegal, Bangladesh’s fragmented early-warning system lacks a unified alerting protocol, which limits its ability to reach remote communities with clear, actionable warnings.
This recognition has shaped the evolution of impact-based forecasting mechanisms, which aim to connect early-warning information to specific impacts on communities, thereby enhancing their preparedness. While impact-based forecasting has been piloted in some areas, it is still in the early stages of implementation and needs scaling to support multi-hazard early-warning systems. Impact-based forecasting requires greater coordination between national and subnational stakeholders, alongside support from regional and global actors, to advance risk-informed communication and preparedness.
