What We Have Here is a Failure to Coordinate:
Lessons Learned Problems Observed in the Response
to the Tsunami

The more than 200,000 deaths from the December 2004 Indian Ocean tsunami made it one of the worst natural disasters of the twentieth century. It is also notable as one of the best-funded and best-attended disasters of all time, with hundreds of aid actors on the ground in Indonesia, Sri Lanka, Thailand, the Maldives, and the other affected areas, and current estimates of the total tsunami donations stretching into the billions of dollars.

The International Federation of Red Cross and Red Crescent Societies (IFRC), in their World Disasters Report 2005, describe the situation among relief agencies, local authorities, and tsunami victims in Aceh as an “information black hole” (IFRC, 2005), noting that

I was on the ground in Indonesia almost immediately after the tsunami1, working as a doctor for the International Rescue Committee (IRC), and so was able to get a firsthand view of relief efforts: for the first week or so from Banda Aceh, in daily attendance of the UN coordination meetings, and then as a doctor in the devastated areas along the west coast of Aceh, with only intermittent communication with, or travel to, IRC headquarters in Banda Aceh. Although my initial view was very focused on my own perceptions, in the months since the tsunami, I’ve been able to glean a broader understanding of others’ perceptions, both NGO and military, through formal and informal discussions of how the relief worked and how it didn’t, with a focus on coordination and information sharing – or the lack of it – among the participants. Here is what we collectively observed on the ground.

The “logistics” mentioned in the IFRC report largely refer to transportation: the destruction of
the main coastal road in Aceh left most civilian aid agencies stuck in Banda Aceh, unable to assess damage (or provide services) farther down the coast.
The aid agencies were largely operating in the dark: they didn’t know how many people required medical care; they didn’t know how many people were without food, or without shelter; they didn’t know how many children were unattended by adults. This made it impossible to efficiently order supplies, or to measure progress: you can’t tell how far you have come if you don’t know where you started.

Even after transportation needs were addressed and more information-gathering became possible, there seemed to be little inclination on the part of organizations to collaborate on information-gathering activities, leading to the unfortunately classic situation in which one group of displaced persons is “assessed” multiple times in the space of a few days – and in which that same group complains that despite all the assessments they have not yet seen any benefit.

The lack of information was compounded by a lack of sharing between the military and civilian components of relief: during the first weeks the situation was extremely chaotic in Banda Aceh with no one visibly in charge. In addition, the civilian agencies did not seem pressed to coordinate with each other or with either the Indonesian or the American military. Having spent years working at the Center of Excellence in Disaster Management and Humanitarian Assistance (COE), an organization which makes a great effort to promote civil-military cooperation in disasters, I found
this lack of initial coordination probably the most dismaying aspect of the relief.

The lack of desire to coordinate with the Indonesian military seemed, in my observation, to be largely due to strong feelings among aid workers concerning the role of the Indonesian military in repression of the population of East Timor in 2002. Another issue was that many of the aid workers were accustomed to working in the context of countries with weak or nonexistent central governments (e.g. Somalia) or states that are actively persecuting the aid recipients (e.g. Sudan): those kinds of experiences do not provide a venue for learning how to coordinate well with local governments, and in this setting some NGO personnel were actively resisting the need to coordinate with the Indonesians.

In one meeting, in particular, I remember relief personnel discussing with resentment and annoyance the Indonesian military’s insistence that aid agencies report their whereabouts and activities at a daily coordination meeting – this although the military was clearly representing the legitimate, and democratically-elected government of Indonesia.

The lack of coordination between civilian relief agencies and the U.S. military was even more of a problem, because U.S. military forces possessed almost all of the air transport assets. In a situation in which the main coastal road down the west coast of Aceh had been destroyed, this meant that if someone wanted to travel down the coast in the initial weeks after the tsunami, it would almost certainly necessitate getting transport from the U.S. military (few organizations except for Médecins Sans Frontières [MSF] had managed to obtain their own helicopters). Despite this, it took 2-3 weeks for organized attempts to connect military transport with civilian aid resources to bear fruit; two to three weeks in which, starved for information and lacking transportation, the aid agencies occupied themselves with setting up bases in Banda Aceh and with having meetings; two to three weeks in which no military liaison was present at those meetings. Two to three weeks in which the needs of the worst-affected populations down the coast went largely unmet (except through the efforts of the Indonesian military, in many places the only relief presence).

The U.S. military, meanwhile, continued to drop water and basic supplies up and down the coast, with apparently little understanding that the civilian relief agencies had a variety of unique and essential resources, personnel, and experience. When I asked one military officer about this, pointing out that the civilian agencies had some resources that the military did not possess, he was incredulous and challenged me for an example. When I suggested sanitation teams and child protection services, off the top of my head, he admitted that these were areas that the military was not providing that might be of use in a devastated area with thousands of refugees.

Making the lack of coordination among the aid agencies even worse was the failure to utilize modern communications and coordination systems for information sharing. There was no visible attempt to provide any kind of widespread Internet access in Banda Aceh for relief agencies: some had it; some did not. This meant no email, no online forums for information discussion, no online bulletin boards for coordination or for posting results of assessments. In short, there was a highly compromised ability to electronically share information across the field of relief actors. Having managed to obtain dialup access myself, I remember suggesting at one health meeting that I could post the results of one needs assessment on a website for others to view and download, and being informed that none of the other attendees had access to the Internet. Only the efforts of one NGO, Télécoms Sans Frontières, were directed towards providing reliable Internet access to all agencies,
and their resources were strained to the limit.
Even once Internet access became common, however, there seemed little use of, or even awareness of, the tools available online specifically for the purpose of sharing information and coordinating activities. As in all of the previous disasters in which I have played a role, I found no evidence that anyone was utilizing either the excellent free online sites such as Backpack (www.backpackit.com) and Yahoo Groups (groups.yahoo.com) or the paid sites providing greater and more secure functionality such as Jotspot (www.jotspot.com), SocialText (www.socialtext.com), and Basecamp (www.basecamp.com). These online collaborative tools are all designed for common editing and viewing of: groups of documents, lists, databases, photos, maps, assessments, etc. — and even the functionality provided by the free sites is a vast improvement over simply sending out multiple emails to multiple recipients. Still, the only common electronic tools in use by the time I left Aceh (5 weeks after the tsunami) were electronic mail and cellular telephones, both of which are excellent for one-on-one communication but are unwieldy, inefficient, or just impossible to use for more advanced collaboration, such as collaborative editing of or access to a document or database by large groups. For the aid agencies, it was as though development of the World Wide Web had stopped right after web-based email was invented.

I’ve described a litany of problems and “mistakes observed” (I won’t use the overly-optimistic term “lessons learned” since there is no evidence that any learning of lessons has taken place). What can we do to improve cooperation, coordination, and communications in disaster settings? Here are three simple suggestions:

Create Dedicated Assessment Teams Across Agencies, and Which Include Local Personnel.

It is not acceptable for multiple agencies to be conducting parallel assessments, often in overlapping areas, when time is of the essence (as occurred in this disaster response). In addition, when the same personnel are required, for example, to be providing medical care and also gathering information it is the information-gathering that typically falls by the wayside. Organizations that routinely respond to disasters should expect to do joint assessments with the other groups onsite, and have personnel dedicated and trained in advance for that purpose.

Create a Dedicated Information Coordination Team Responsible Both for Providing Internet Access and Facilitating Use of Web-Based Coordination Tools.

Groups like Télécoms Sans Frontières provide an excellent example of the benefit of having a dedicated group providing access to the Internet. Just having access, however, is not enough: someone needs to facilitate information sharing by creating easy-to-use online spaces to post lists, store files, and keep contact information – all tasks that are required in every disaster setting but which are usually done haphazardly if at all. As mentioned above, there are numerous free tools available that would do the job, but the average responder doesn’t know how to use them – nor should they. A specialist in these essential technologies needs to be present at every disaster site.

Recognize That Civil-Military Liaisons are Absolutely Essential in Settings Where the Military Will Be Playing a Role.

One of my biggest frustrations during the first weeks after the tsunami was not just that no military liaison officer from the U.S. military was present at the UN coordination meetings in Banda Aceh, but also that no one seemed to realize what a huge problem this was – even in a situation where effectively all transport was controlled by the U.S. military and where the aid agencies were unable to travel to where the affected populations were. When no liaison is present, coordinating officials need to push the issue until one is provided (and they need to understand in advance, or find out very quickly, with whom to push the issue).

Much of the work that needs to be done in improving coordination requires that aid organizations actually want to coordinate, and see a value in doing so. Unfortunately, however, there are strong incentives for organizations to want to “fly the flag”. As one local NGO head told me in Aceh “we don’t get donations, time on CNN, or points from the home office, for quietly working in coordination with the other NGOs”. These kinds of incentives produce the kind of chaos and competition seen in so many disaster settings, and require some kind of counterbalancing incentives to compel organizations to coordinate. I don’t know whether incentives should be legal – for example, creating international civil penalties for operating in a disaster zone without authorization from the central authority – or perhaps moral – such as simply publishing a list of the worst offenders – but I do know that it is absolutely essential if we are to stop being part of the disaster ourselves.

References
IFRC. (2005). World Disasters Report 2005 [Chapter 4]. Retrieved from http://www.ifrc.org/publicat/wdr2005/chapter4.asp.

Endnote
1) I arrived January 3rd, eight days after the event.