Skip to content
The case for self-recovery

Along with over half a million Filipino families, Erica and John Rey and their twelve children saw their house blown away by Typhoon Haiyan. Two years on, they had designed and rebuilt their house. They received some materials, a small amount of cash and technical assistance from an international NGO working with a local partner, but essentially the control, the decision making, the day-to-day building were all their own. They are ‘self-recovering’. International and national aid agencies tend to reach between 10 and 20% of those whose housing has been damaged or destroyed in a major disaster.[1] By implication, 80-90% self-recover. With little or no outside support, these families will, in most cases, rebuild their houses with the same vulnerabilities and bad building practices that had been contributory factors to the damage, economic loss, injury or death.

Increasingly, the shelter sector has become aware that the conventional approach to post-disaster housing recovery is only partially fit for purpose. This approach is characterised by an emergency phase typically of three months, followed by early recovery, and then reconstruction. The adoption of transitional or temporary shelters has been popular in recent emergencies – one-room dwellings which are only intended to bridge a gap of a few years until the family rebuilds a permanent house but which frequently exhaust the aid budget and consequently become sub-standard long-term homes. Self-recovery programmes, by contrast, propose that all support should be directed towards the ultimate aim of a safer, permanent house. Although the process is currently little understood, the shelter sector is beginning to develop effective and appropriate approaches to supporting the practice.

The Philippines after Typhoon Haiyan highlights many of the advantages, and some of the pitfalls, of a self-recovery programme. First among the positives is the control, agency and choice exercised by the family. The most striking visual evidence of the sense of ownership stimulated by self-recovery is the delightfully designed woven bamboo screens and the bewildering array of potted plants that adorn the front of every house. A self-recovery approach also reconciles choice and control with an emphasis on safer building methods, through training and community accompaniment. By training local builders in ‘build back safer’ techniques, there is a legacy of better building practice and long-term disaster risk reduction. Each family builds according to its needs (and also, of course, according to its means – a potential negative). By contrast, the off-the-shelf, one-size-fits-all approach of many conventional programmes struggles to assimilate this diversity of need.

Furthermore, many rural Filipino families supplement their income with tiny general stores that are easily incorporated into these self-built, self-designed homes. Thus control over design makes an important contribution to the recovery of livelihoods. Self-recovery has the potential to be quick and effective: a very large number of families can be reached swiftly with cash, some materials and technical assistance and training. As the quantity of cash is much less than the actual cost of the house, the budget can reach many more households than a conventional complete-house programme. Investment in self-build housing in Pakistan after the 2010 floods was found to be comparable, per house, to the cost of a tent.[2]

The Philippines experience also demonstrates some of the challenges confronting a self-recovery programme. Inconsistent technical quality is the most evident. ‘Building back safer’ was in this case promoted through four simple construction messages but compliance was variable. Beneficiary families were subject to a selection process that inevitably excluded a significant portion of the population. The legacy of better building practice therefore failed to permeate throughout the community: many families rebuilt with no improvements to safety. While shelter and livelihoods were integrated, water and sanitation were not; the opportunities to harvest rainwater from new metal roofs and to further the government’s ‘zero open defecation’ campaign were missed.

What works well in the Philippines may not work elsewhere. It is also certainly true that the widespread practice of community self-help in the Philippines and fairly ready access to markets create good conditions for a self-recovery approach. Nonetheless, recent disasters caused by storms, earthquakes and floods in a variety of contexts have shown that a self-recovery approach is often appropriate. Communities are never passive, and the initiation of reconstruction and recovery is an inevitable process.

In 2015, Cyclone Pam devastated the southern islands of the Pacific nation of Vanuatu. Some villages lost almost every house. Within days the families were salvaging material, drying out the palm thatch and beginning to rebuild their homes. There are no markets and few roads on the most affected island of Tanna, and so it was clear from the outset that cash would not be appropriate. The houses there are made almost entirely from natural materials gathered from the nearby forests. Despite considerable logistical challenges, a training programme coupled with distribution of a fixing kit (nails and cyclone strapping) began within a few weeks to support the self-recovery process.

By contrast, self-recovery after the 2015 Gorkha earthquake in Nepal was much less in evidence. A number of factors influenced both the self-recovery process itself and organisational support to it: the stone construction of the housing, the logistics of the mountainous terrain, the delayed delivery of government subsidies, and the need to comply with building codes and standards.

Self-recovery for displaced populations

Research into self-recovery after disasters is fairly recent and has largely concentrated on rural communities that have been affected by ‘natural’ events such as storms, earthquakes and flooding. On the whole these families were not displaced, although disasters are also a cause of forced migration. There are clear differences between the family that has lost its house in a storm and the refugee or IDP family fleeing their home. The former can build back on its own land; the latter may have to settle in a camp or precariously on the edge of a city. Are there, nonetheless, potential benefits from looking at shelter solutions for displaced and migrant populations through the lens of self-recovery?

Only some 30% of refugees and IDPs across the world are housed by international organisations. The remaining 70% are in rented accommodation, hosted by friends and family, sleeping rough or in home-made makeshift shelters. In one way or another they are ‘self-recovering’ – if by that we mean the process of recovery, or coping, using the family’s own resources, without significant outside intervention and with considerable control over their pathway to recovery. There are many examples of refugees, IDPs and disaster-displaced people – particularly in urban settings – who have by this definition self-recovered.

Refugees from the former Spanish colony of Western Sahara have lived in camps in Algeria since 1976. The harsh desert climate and their nomadic heritage demand very specific solutions to their housing needs. With temperatures reaching 50°C during the day but it being cold at night, the Sahrawi have two dwellings: a large green tent and a mud-brick house with windows close to the ground for cool ventilation. The solidity and thermal mass of the mud bricks and the well-ventilated airy tent provide a suitable combination of environments. International NGOs and solidarity groups provide canvas for the tents but the design, fabrication and siting are entirely controlled by the refugees. In terms of control, choice and agency – at least with respect to their housing – they are self-recovering.

In the Kakuma refugee camp, opened in north-west Kenya in 1992, many residents have developed a deep sense of pride in their houses. They have decorated them to their individual tastes and values, and planted trees and flowers outside, even engaging in competition with neighbours over the personalisation of living spaces.[3] Such actions demonstrate the ways in which self-recovery, as understood here, also occurs in situations of protracted displacement.

It is evident that self-recovery in the context of a natural disaster is a spontaneous process. It is also evident that many displaced populations too exercise choice and agency over their housing options. Frequently refugees and IDPs are left with no other choice than to cope on their own. Both the potential for self-recovery programmes and the benefits of an approach that has more to do with empowerment than actual construction might be relevant – although undeniably different in nuance, detail and context from post-disaster situations – to the circumstances of refugees, IDPs and those displaced by disaster.


Bill Flinn
Senior Shelter Advisor, CARE International UK, and manager of the research programme ‘Promoting Safer Building: supporting self-recovery’

Holly Schofield
Researcher, CARE International UK

Luisa Miranda Morel
Research assistant, CARE International UK

This research is a collaboration between CARE UK, the Overseas Development Institute, University College London and the British Geological Survey, funded by the Global Challenges Research Fund.

[1] Parrack C, Flinn B, Passey M (2014) ‘Getting the Message Across for Safer Self-Recovery in Post-Disaster Shelter’, Open House International 39 (3)

[2] Hendriks E, Basso M, Sposini D, van Ewijk L, Jurkowska H (2016) ‘Self-built housing as an alternative for post-disaster recovery’. ETH Zurich

[3] Feyissa Demo A (2009) ‘Riding on the back of a tortoise’ Forced Migration Review issue 33


This site is registered on as a development site. Switch to a production site key to remove this banner.