Changing approaches to post-disaster shelter

For decades, development and humanitarian architects have stressed the importance of an enabling approach to reconstruction that recognises the central role that affected people play in rebuilding their homes in the wake of disaster. Yet this rhetoric has seldom translated into action, and shelter responses are typified by the provision of inadequate, inappropriate and badly built shelters to a small proportion of the affected population. The success of a shelter response tends to be measured by the number of units provided, and there is pressure to help as many people as possible as quickly as possible. Agencies are prompted to provide larger quantities of smaller, cheaper, temporary houses, at the expense of quality and durability.

This article argues for a shift in the international community’s approach to post-disaster housing reconstruction. It is an argument for less physical building and a greater concentration on helping affected people rebuild their homes themselves. Two central and connected themes support this argument: the growing concern to incorporate disaster risk reduction (DRR) into the rebuilding process; and the sheer scale of the challenge of rebuilding.

Reducing disaster risk

“There is an increasing acknowledgment that DRR is a major requirement for coping with emergencies and enhancing community resilience. It makes more sense to protect communities from disasters ahead of time than to wait for them to happen before responding.” Save the Children UK, At the Crossroads: Humanitarianism for the Next Decade, 2010, p. 17.

This Save the Children analysis from 2010 points out the benefits of closing the door before the horse has bolted (for every dollar spent on DRR, seven dollars of losses can be prevented). However, it is difficult to predict where and when the next disaster will strike, and not easy to raise funds for a disaster that has not yet happened. Nonetheless, there is country-level evidence that risk reduction and preparedness can save lives and reduce the cost of reconstruction. In Cuba, for instance, a national focus on DRR has drastically reduced loss of life from hurricanes; while Hurricane Katrina killed 1,800 people in the United States in 2005 and Cyclone Nargis killed 146,000 people in Myanmar in 2008, three hurricanes that crossed Cuba during the 2008 season killed a total of seven people. For more on Cuba’s approach to DRR, see Oxfam America, Weathering the Storm, 2004.  In Bangladesh, cyclone fatalities have declined dramatically over the past 40 years through disaster preparedness measures, the mobilisation of the Bangladeshi Red Crescent and the construction of community cyclone shelters.

While it is often difficult to raise money for preparedness, disasters do attract funding and there is a prevailing view that the opportunity disasters present for measures to reduce vulnerability and risk in the future should not be missed. Yet post-disaster efforts to reduce vulnerability are held back by the understandable rush to respond. There is often a failure to recognise that effective DRR demands a steady, slow approach. After a disaster, efforts to get messages across about reducing risk through safer building methods or materials are often limited to rapidly produced posters and leaflets, despite commitments under the 2005 Hyogo Framework for Action to include DRR as part of any post-disaster response.

Scale: the numbers defy imagination

Disasters are never the same and the international community’s response differs widely depending on a whole gamut of circumstances. Small-scale rural emergencies may be adequately addressed without the involvement of the international community at all. In contrast, well-funded major urban disasters in the media spotlight attract huge numbers of international NGOs. No two disasters are the same.

Running the risk of generalisation, in most disasters with large shelter components the coverage achieved by the relief effort falls woefully short of the scale of need. The amount raised for the Indian Ocean tsunami far outstripped any disaster before or since and was unique in providing enough funds to meet housing needs. Normally, only 10– 20% of housing needs are met, frequently with temporary rather than more permanent housing. To cite a few examples: one year after Cyclone Sidr the number of dwellings built by aid agencies in Bangladesh (2007) represented 7% of need; in Padang, Indonesia, after the 2009 earthquake it was 14%; and in the Pakistan floods (2010) it was 2.5%. Eighteen months after the earthquake in Haiti around a third of housing needs had been met. In all of these cases the dwellings constructed were ‘transitional’ shelters intended to bridge the gap between a tent and a permanent house. While the transitional shelter can often serve a vital function, it has also been criticised as wasteful of materials and frequently becomes post facto permanent. DEC, Urban Disasters – Lessons from Haiti, 2011.

These are percentage figures. But when the total numbers are taken into consideration, it becomes apparent that the size of the task is quite beyond the capacity of aid agencies. Cyclone Nargis in Myanmar (2009) destroyed 450,000 houses, Cyclone Sidr in Bangladesh (2007) 400,000, the Pakistan earthquake (2005) 400,000 and the floods in Pakistan in 2010 an extraordinary 1.6 million homes. To put this into perspective, the number of houses built in England by the entire construction industry is typically in the order of 150,000 a year. Very few agencies that engage in shelter after disaster have departments or staff dedicated to construction. Nearly all rely on external consultants drawn from a very small pool.

It is hard to draw any other conclusion: the aid community clearly does not have the capacity to provide post-disaster housing on a significant scale, and this must be the responsibility of national governments and their donors.

Recognising this, NGOs try to concentrate on the most vulnerable – the old, the infirm, the disabled, womenheaded families, families with many children. Yet in practice these vulnerable groups, including the homeless and landless, tend to be overlooked by shelter programmes. In Bangladesh it proved extremely difficult to build even temporary shelters for families illegally squatting on stateowned land; in Haiti little could be done for tenants or for families living illegally on steep ravine slopes. It is much easier to provide transitional houses for families that own their land and have a bit of space free of rubble. These are almost certainly not the most needy.

A different approach to shelter after disaster

People made homeless by a natural disaster ‘are the first responders during an emergency and the most critical partners in reconstruction’. World Bank and GFDRR, Safer Homes, Stronger Communities, 2010.  There are few principles in the world of shelter that can be said to be universally true, but perhaps this is one that most will agree on. In the jargon of the aid world, this is the community’s social capital and its greatest asset. However, this point is frequently overlooked in the quest for a product and the measurement of impact by the length of the beneficiary list.

If the international community can only meet 20% of the need, then we can only conclude that 80% is met by affected people themselves, rebuilding their homes using their own labour or hiring the local builder. Once again, Haiti provides graphic illustration: 18 months after the earthquake some 50,000 permanent houses had been built using Haitians’ own resources. The international community had no influence on the quality of the construction. Houses are being rebuilt incorporating preearthquake weaknesses. Haiti also demonstrates another phenomenon that will be increasingly dominant: the money that flowed in from the diaspora in the form of remittances outstripped the total value of aid money several-fold. Haiti is being rebuilt by Haitians.

With disasters on the increase See Roger Musson, The Million Death Quake (Basingstoke: Macmillan, 2012).  and funds limited, it is inevitable that most families will rebuild their homes using their own resources. This is the challenge that the shelter sector has to step up to. How, in the aftermath of a major disaster, do we intervene to ensure that safer and more durable – yet still affordable – houses are built? One approach might be training in building for safety. This does not imply a total change of emphasis away from construction and towards a training-only approach: many of the current approaches to shelter after disaster – cash transfers, the provision of materials, demonstration projects – can incorporate a training element that ensures that safer building practices permeate the community. Currently there is a startling lack of engineering and construction capacity in the sector – something that must be addressed to increase quality and to ensure that we build back better and more safely.

Hard decisions have to be made to balance the imperative to meet immediate needs with the long-term benefits of a more developmental approach that promotes a culture of safer building. This article argues for less building and more facilitation of good, safe building practice in the knowledge that affected families, immediately or once the transitional shelter has fallen apart, will ultimately take responsibility for the safe building of their new home.

Bill Flinn is Associate Lecturer at the Centre for Development and Emergency Practice (CENDEP), Oxford Brookes University.