The first mile of warning systems: who’s sharing what with whom?

November 8, 2018
JC Gaillard and Ilan Kelman
Local youth are setting up a localised warning system for flooding.

Inclusive warning systems

Warning systems for hazards used to be assumed to be top-down: supply technology, data and messages, and then connect to the people affected as the ‘last mile’ of the warning system. Yet lessons from past decades W. A. Anderson, ‘Disaster warning and communication processes in two communities’, Journal of Communication 19, 1969; E. C. Gruntfest, T. E. Downing and G. F. White, ‘Big Thompson flood exposes need for better flood reaction system to save lives’, Civil Engineering 48(2), 1978. alongside recent work M. H. Glantz et al., ‘What does it mean to be El Niño ready?’, Atmosphere 9(3), 2018 (; Z. Zommers and A. Singh (eds), Reducing disaster: early warning systems for climate change (Dordrecht: Springer, 2014). explain why bringing in affected people last creates problems. Instead, warning systems need to be inclusive from the beginning.

Inclusion is a tricky concept. It is easier to articulate in policy guidelines than to translate into and apply in practice. Inclusion entails sharing power to benefit people who usually lack the opportunities to make decisions affecting their everyday lives. This includes having the appropriate warning information, preparedness and understanding to make informed choices long before a hazard manifests, and when faced with impending and potentially harmful events or processes. Inclusion is therefore a political process, and as such can be resisted by the scientists, governments and agencies that have long been the key providers of warning information and systems.

Why should warning be inclusive?

Warning of potentially harmful events and processes has long been a matter of providing external and usually scientific knowledge to people who may be affected (Figure 1). It has been a largely technocratic and top-down approach handled by scientists and filtered by governmental and non-governmental agencies and officials. B. Wisner, J. C. Gaillard and I. Kelman (eds), Handbook of hazards and disaster risk reduction (Abingdon: Routledge, 2012). Especially since the 2004 Indian Ocean tsunami, the main challenge has been improving the ‘last mile’ of the chain of actions designed to reach people on the ground. F. Thomalla et al., From knowledge to action: learning to go the last mile (Stockholm: Stockholm Environment Institute, 2009). Even so, there is still an assumption that the signal has to come from the top-level experts down to the (apparently ignorant) people affected.

Figure 1: Signs indicating tsunami-prone areas and evacuation routes in Iquique, Chile
Figure 1: Signs indicating tsunami-prone areas and evacuation routes in Iquique, Chile

Instead, inspired by development work, R. Chambers, Rural development: putting the last first (London: Longman, 1983). it has been argued that the last mile should rather be the first, transferring leadership to local people in designing, operating and responding to warning systems. I. Kelman and M. H. Glantz, ‘Early warning systems defined’, in Z. Zommers and A. Singh (eds), Reducing disaster: early warning systems for climate change (Dordrecht: Springer, 2014). Fostering and supporting the inclusion of local people, including the most marginalised, in warning systems requires recognising that they have some level of knowledge of local contexts and hazards, and that their concerns and priorities guide how they deal with distant and impending hazards. The people who need warnings are not helpless or passive victims: they display a wide range of capacities often including the skills, knowledge and ability to act, and the resources to do so. Wisner, Gaillard and Kelman, Handbook of hazards.

Inclusive warning: challenges and opportunities

The main challenge is ensuring that the last and first miles come together to provide an inclusive and integrated approach to warning based on everyone’s knowledge and concerns. Neither top-down nor bottom-up should dominate or be ignored. Such integration is, however, difficult, particularly for local, sometimes indigenous or traditional, knowledge and scientific and/or external knowledge, both of which are valuable in designing and operating warning systems. J. Mercer et al., ‘Framework for integrating indigenous and scientific knowledge for disaster risk reduction’, Disasters 34(1), 2010. Each relies on different worldviews and epistemologies which might be difficult to reconcile. Co-existence and connection, rather than integration, of different forms of knowledge might be more successful. This entails granting access to both local, including indigenous and traditional, and scientific and/or external knowledge to everyone in order to make informed, collaborative decisions based on the largest possible array of tangible and trusted information.

Additionally, local people (like scientists, governments and agencies) are never a homogenous group of individuals sharing, requiring or trusting the same form of warning-related material and information. They have vulnerabilities and capacities reflecting their own positions within society and their environments. Wisner, Gaillard and Kelman, Handbook of hazards. It is therefore essential to consider local people in their diversity, especially those who are at the margins, such as older people, prisoners, people with disabilities, homeless people, children and minorities.

In many societies, fostering the inclusion of the most marginalised people in designing and implementing warning systems may challenge existing cultural norms and values. J. C. Gaillard and M. Fordham, ‘Silent, silenced and less-heard voices in disaster risk reduction: challenges and opportunities towards inclusion’, Australian Journal of Emergency Management 33(1), forthcoming. This potential tension between the moral imperative to address the concerns of the most vulnerable and respecting traditions is complex and context-specific. A balance needs to be struck between respect for local culture and local people’s inclusion, while recognising that some cultural and legal changes may be necessary.

It is not enough to work with only one specific group of people such as older individuals, indigenous people, or children. This would not address inequitable power relations (or their inability to make informed decisions by themselves) that underpin their vulnerabilities and prevent recognition of their capacities. Inclusive warning should be a process through which scientists, governments, and agencies recognise and accept the specific vulnerabilities and capacities of different local groups. Meanwhile, local people should be able to trust and work with information and advice issued from external sources.

Warning for typhoons and floods in a Philippine jail

Prisoners are a good example of minority groups who frequently fall between the cracks of classic last-mile warning systems. The last mile often fails to overcome spatial marginalisation and social neglect, so many inmates often suffer silently in disasters. J. C. Gaillard and F. Navizet, ‘Prisons, prisoners and disaster’, International Journal of Disaster Risk Reduction 1(1), 2012.

The San Mateo Municipal Jail in the Philippines is a small, old detention facility housing more than 550 male and female inmates, spread across five overcrowded cells. Following a hasty and perilous evacuation of the jail during Typhoon Ondoy in 2009, cell wardens designed a sophisticated warning and evacuation plan for storms and other natural hazards. The plan was designed in collaboration with the inmates and national and local government agencies, with support from civil society organisations and scientists.

When a typhoon is about to strike San Mateo, wardens and guards monitor the water level in the adjacent Marikina River and coordinate with government agencies to organise vehicles and shelters for a possible evacuation. Meanwhile, inmates get ready in their cells. They organise themselves by predefined small groups, comprising pairs of individuals (usually one younger and one older prisoner), under the command of inmate leaders. Each group carries a ‘go box’ with key items such as a torch, biscuits and drinking water.

Inmate leaders ensure that their fellow inmates take their sleeping mat and linen with them. When the wardens and government officials order the evacuation, buses pick up the inmates at the prison gate under the supervision of local guards, reinforced by local police and personnel from the government’s regional headquarters. On arrival, inmates organise the evacuation centre, tidy it up when the emergency ends and take the lead in cleaning the jail when they return.

This warning and evacuation process was successfully implemented when Typhoon Ompong struck the Philippines in September 2018 (Figure 2). The jail was safely evacuated before flooding badly affected the facility. Prior multiple drills made the process smooth and straightforward. Relatives of the inmates, who also participated in the drills, were kept informed of the evacuation procedure.

Figure 2: Evacuation of San Mateo Municipal Jail in the Philippines at the onset of Typhoon Ompong in September 2018
Figure 2: Evacuation of San Mateo Municipal Jail in the Philippines at the onset of Typhoon Ompong in September 2018

Fostering inclusion in warning systems

The San Mateo example shows that bridging the gap between the last and the first mile is possible. It demonstrates that involving everyone is essential to:

  1. Ensure that everyone recognises each other’s vulnerabilities and capacities.
  2. Build rapport and trust so that everyone’s potential contribution is recognised and included in designing and operating inclusive warning systems.
  3. Make inclusion in warning systems culturally relevant and acceptable to all sectors of society.

Exchanges among local people dealing with natural hazards and those with usually more power in producing and running warning systems – including scientists, governments and civil society organisations – requires appropriate social and political space. The process must allow for conversations to occur without necessarily breaking laws or local cultural norms and values. It must also accept that laws, cultures and values which engrain discrimination, oppression and other social ills might need to be reconsidered and possibly altered, in tandem with the people affected, before any first mile warning system can be truly successful. Frequently, it will mean challenging how warning systems and wider disaster-related initiatives are framed and structured, and accepting them as people-led processes. This often long and winding road starts with the first mile. In the end, this journey will lead to warning systems that involve local people from the beginning. Whether for the next tsunami in Indonesia or the next hurricane in the US, the first mile of warning systems is about involving and serving those directly affected.

JC Gaillard is Associate Professor in the School of Environment at the University of Auckland. Ilan Kelman is a Reader in Risk, Resilience and Global Health at the Institute for Risk and Disaster Reduction and the Institute for Global Health at University College London, and a Professor II at the University of Agder in Norway.

An earlier version of this article was presented at the 12th APEC Senior Disaster Management Officials Forum in Kokopo, Papua New Guinea, on 25–26 September 2018.


Comments are available for logged in members only.