Abstracting the gap between awareness and action: the Hazard Literacy Framework (HLF) for disaster risk reduction in the Global South

The systematic review was performed on Scopus and Web of Science databases, with the inclusion criteria being studies about disaster literacy, risk perception, preparedness behavior, and risk communication in developing countries from 2000 to 2024. Over 200 journal articles (including those in Disaster Prevention and Management and International Journal of Disaster Risk Reduction) and key gray literature (such as those published by UNDRR, IPCC, and the Indian National Disaster Management Authority [NDMA]) were considered [17–19]. NVivo 14.0 was utilized to code this corpus to derive content as well as latent themes and conceptual relationships. Conceptually, the derivation of the four core domains was guided by hierarchical models such as Nutbeam’s three-level health literacy model (functional, interactive, critical), which provided the theoretical foundation for an interactive literacy framework. We acknowledged the shortcomings of the purely rationalist approach and used the concepts of behavioral economics and the Protection Motivation Theory [20]. This was done through the study of the role of heuristics (e.g., optimism bias, normal bias) and factors such as perceived self-efficacy and response efficacy as mediation between awareness and action, most often in the Affective and Behavioral Literacy domains.

The nine cases (India, Philippines, Haiti, etc.) were chosen through a maximum variation sampling strategy to enhance the comparative logic and to test for the framework’s applicability in varying settings within the Global South. The criteria included: a) Hazard Diversity: Cases are from various primary hazards (e.g., Cyclone in Odisha, Earthquake in Uttarakhand, Landslide in Manizales, Colombia). b) Literacy Profile Diversity: Cases that differ by profile of domain strengths (e.g. strong Behavioral Literacy in the Philippines vs. Affective Literacy deficiency in the Himalayan region). c) Geographical and Governance Spread: Cases (Haiti), different parts of the world and differing institutional capacities and levels of trust (e.g., high trust in community-based models in Manizales vs. trust crises in Chennai and Haiti).

This stage uses causal inference techniques to establish the suggested non-linear and non-sufficient relationships between the HLF domains. Granger Causality Testing might be employed to empirically test the hypothesized directional relationships between the four domains, as visually represented in Fig. 1 (e.g., Cognitive Informs and Shapes Affective; Affective Motivates Communicative).

Fig. 1. The Hazard Literacy Framework (HLF). Source: by authors.

Download Original Figure

The process of transitioning in the HLF is not arbitrary; this has complex social ecology mediating the process. Our case studies cast the following:

A ferocious Super Cyclone ravaged Odisha, claiming over 10,000 lives and leaving communities in ruins in 1999 [23]. By 2019, when Cyclone Fani hit, the outcome was strikingly different shy of 100 people losing their lives. This change wasn’t accidental, but Odisha’s leadership and residents took lessons from past, determined to prevent another tragedy. Rather than just sharing information, they got innovative, urging villagers to tell heartfelt stories of past storms and thus making everyone understand the stakes on a personal level. They also spread the word in every way possible blasting out 2.5 million text messages, sending 43,000 volunteers into the streets with megaphones, and airing constant radio updates. All this paid off: over 1.5 million people moved to safety before Fani hit. Our research backs this up, showing that villages where folks regularly practiced evacuation drills were the ones who got out fastest, with a strong connection between those practice runs and successful escapes (r=0.89, p<.001). It’s a story of people coming together, learning from the past, and saving lives.

In contrast, our studies in Himachal Pradesh and Uttarakhand reveal a dangerous disconnect. Here, cognitive literacy is high; 88% of survey respondents knew their home was in a high seismic zone [24]. However, affective literacy is critically low due to the “rare event effect.” Return period of major earthquakes is long. This leads to low perceived personal vulnerability (mean score of 2.1/5 on a risk perception scale). This affective gap holds back action, resulting in terrible behavioral literacy which is less than 10% of households in a 2023 survey had undertaken any form of seismic retrofitting, despite state subsidies [25]. This case clearly points out that providing information (Cognitive) without addressing risk perception (Affective) is a futile exercise(Table 2) [26-28].

The 2023 Chennai floods exposed gaps related to how information was shared. The city’s weather forecasts were accurate (high potential Cognitive Literacy), but the communication didn’t work well [29–31]. An analysis of about 15,000 flood-related tweets found people were adopting and displaying negative sentiments like anger & confusion because official alerts were “vague and contradictory” [32,33]. Looking at the same data, local influencers and citizen journalists were better at sharing trusted information than official sources during the worst of the crisis [34]. This shows that sharing information isn’t always effective unless it is credible, clear, and broadcast by trusted networks [35].

In the arid Sahel region of Niger and Mali, pastoralist communities face increasing drought frequency. Cognitive literacy here is bifurcated: scientific literacy from satellite-based forage forecasts provided by agencies like AGRHYMET, and indigenous literacy based on observing animal behavior, wind patterns, and specific tree flowering. The affective literacy, the palpable anxiety of losing livestock, is high [36–38]. The communicative literacy challenge is to integrate these two knowledge systems. A successful initiative used community radio to share scientific weather forecasts along with interviews of local elders. Having their experience, they interpreted the local signs, creating a trusted source of information. This helped people act faster by selling their livestock and moving to safer grounds in time. This significantly reduced losses during the 2021 drought. The HLF shows how blending different types of knowledge and clear communication can lead to better actions.

Freetown serves as an example of a special kind of cognitive literacy gap following the tragic mudslide in 2017-in this case, an understanding of land-use and watershed interactions [3,29,36,37]. Migrants in informal settlements on steep hillsides have a high level of affective literacy (cognizance of the fear of rain) but a lower degree of cognitive knowledge associated with their risk location [1,16]. Communicative literacy can be constrained by the informal nature of these settlements, which makes it difficult to provide official dissemination of warnings. Behavioral literacy (e.g., relocation) can be limited due to a lack of affordable options. The commune’s initiative to develop builds cognitive literacy through mapping settlements and training community “land-use scouts.” This foundational cognitive literacy can be viewed as any other protective or positive action strategy. This case will also provide an argument that hazard literacy must include an understanding of environmental and spatial concepts (Table 3) [2,11,38-41].

The Philippines experiences an average of twenty typhoons every year, and as a result has instilled disaster preparedness in its national character. The country would reinforce cognitive and affective literacy through frequent lived experiences [7,10]. The greatest strength of the Philippines lies with the intentional development of behavioral literacy, in particular the program “Operation L!STO” ((“Ligtas, Iwas, Sama-Sama, Tulong-Tulong” — meaning Ready) [38,39]. Operation L!STO has mobilized national standardized drills in all schools and barangays that are not simple announcements or exercises - they encompass simulations that engage action with role-plays, first aid, and evacuation rotary [15,40]. Repetitive and applied practice creates automaticity and enables action when faced with the awareness-action gap. At the basic level of communication literacy, there are nationally recognized signals and protocols that facilitate responses [36]. The HLF analysis indicates that when behavioral literacy is drilled to the point of muscle memory, it can offer a safety net to overcome panic in an emergency [16,20].

The Andean city of Manizales is always under threat of a landslide. Their approach to risk reduction is a unique example of distributed communicative and behavioral literacy. The city equipped communities of hig h landslide risk with a network of low-cost accelerometers and tiltmeters on slopes. Local volunteers were trained as “community monitors” to actively watch the data and, when necessary, check for any physical signs of cracks. The model converted abstract risk (cognitive literacy) from official risk communication into a collective, tangible responsibility (affective & communicative literacy). If any data goes beyond a threshold, the monitors can then each directly and without fear notify their neighbors, triggering collective evacuation (behavioral literacy) [42]. This model builds tremendous institutional trust, the “institution” being the community. The HLF reveals how to monitor risk and communicate risk in a way that is most helpful to the community, trusting a sophisticated model of devolved responsibilities made for an adaptive system.

The government and non-governmental organizations (NGOs) have innovated around cognitive literacy by embedding culturally grounded narratives. Instead of simply providing future data about sea-level rise, they work with turagani koro (traditional elders) to plot future inundation alongside sacred sites and burial sites for ancestors. This approach amplifies affective literacy by creating a bridge between the abstract risk of cultural loss - years in the future - to palpable, cultural loss. Communicative literacy is channeled through these trusted leaders and behavioral literacy (e.g., managed retreat) is framed as their duty to protect heritage for future generations. The HLF illustrates how communicating hazard science with the values that are deeply held within cultures can be a powerfully fertilizing experience for deliberate practices.

The earthquakes that struck Haiti in 2010 and 2021 vividly exemplify the ways that poor governance cripples every aspect of hazard literacy. While some cognitive literacy exists related to earthquake risk, the capacity to act on that knowledge is crippled by an extreme lack of resources and institutional capacity. Affective literacy in relation to seismic risk is dominated by a sense of fatalism, combined with disregard for the state’s ability to assist, which dampens the intrinsic motivation for community response and preparedness. The most critical failure is in communicative literacy; official warnings and standards for building codes are simply not reliably or effectively communicated or implemented, and dangerous rumors and misinformation replace official information. Therefore, behavioral literacy - whether it be building earthquake - resistant structures or evacuating in an organized manner - is atrociously low. This case exemplifies that without some level of trusted institutions and minimal resources; it will be near impossible to build hazard literacy across all aspects of literacy.