Over the last 15 years CARE India and other NGOs have repeatedly responded to natural disasters where large numbers of people have lost their homes. These responses have frequently included both provision of short-term emergency shelter and construction of more durable housing, often designated transitional or permanent.
While there have been individual evaluations of some of CARE’s programmes immediately upon completion, there has not been a comprehensive study of the medium- and long-term outcomes of post-disaster shelter programmes undertaken by CARE or many of its peer agencies. This study aims to evaluate the medium- to long-term effectiveness of postdisaster shelter responses and recommend measures to strengthen future shelter programmes, whether undertaken by CARE or other agencies, to most effectively address the complex and interconnected needs of disaster-affected women, girls, men & boys.
The study covers 10 shelter programmes undertaken after natural disasters by CARE since 2001 and furthermore draws on separate evaluations of 3 Christian Aid & SEEDS post-disaster shelter programmes. The study covers projects in 12 States & Union Territory (Andaman & Nicobar Islands, Andhra Pradesh, Bihar, Gujarat, Jammu & Kashmir, Odisha, Rajasthan, Tamil Nadu, Telangana,
Uttar Pradesh, Uttarakhand & West Bengal) and includes projects undertaken after the Indian Ocean tsunami, earthquakes, floods and cyclones.
The construction of durable houses as part of postdisaster shelter recovery programmes delivers both essential safe and dignified shelter, and a valuable asset, to the beneficiaries. This gives beneficiaries the security to focus on other urgent priorities and prevents them falling into destitution. Projects have generally increased the robustness of houses and successfully reduced risk of future natural disasters. There are several examples of houses built in the projects studies surviving significant natural hazards.
Post-disaster shelter programmes which provide durable housing successfully meet humanitarian needs and protect the vulnerable after disasters.
Delivering shelter recovery programmes is complex and often subject to significant competing interests and obstacles. The needs of women, girls, men and boys, and the needs of different households, can vary significantly. A one-size-fits-all shelter design has limited flexibility to meet these varied needs.
Generally projects have focussed mainly on the shelter product to be delivered and not enough on building capacity and agency of the beneficiaries. The durability of shelter is a critical component of the longer-term success of shelter recovery programmes. Maintenance burden and costs, and the economic capacity of beneficiaries, are key drivers for, or obstacles to, good longterm outcomes of shelter programmes:
• Those who can mobilise the economic resources have built upon the asset they have been given (often literally) to make their house provide for all their needs, including the specific needs of women, girls, men and boys, and often to grow their income. The shelter assistance they have received has both protected them and given them the opportunity to improve their lives and reduce their poverty.
• Those who cannot mobilise economic resources – the very poorest, most vulnerable people in society – have been unable use their housing in this way. Secure shelter has allowed them to use their economic resources to survive, has protected them and met their urgent needs, but it has not led to a reduction in their poverty and the risks and vulnerabilities that come with this. They remain trapped in what many consider unsuitable housing which provides basic shelter but not much more. The specific needs of women, girls, men and boys in households remain unmet.
The projects studied have a limited range of approaches to delivering shelter assistance, essentially contractor-built durable, pukka houses or contractor-built houses with a mixture of durable and temporary materials. Importantly, designs used always considered local construction practice and used local materials and were appropriate to the context and local hazards. This improved acceptance, ownership and ability to maximise the value of shelters, and contributed significantly to effective disaster risk reduction, and should be encouraged in any future projects.
The scale and reach of the projects studied varied significantly. Some projects met a significant portion of the need in the context of a disaster where with many actors coverage of assistance was very good. Others met a very small proportion of the need in a context where there were no other actors. Inevitably in the projects with fewer resources the value of assistance must be reduced and the resources must be focussed on the most vulnerable. Projects did this by targeting both geographically and based on vulnerability (Schedule Castes, Schedule Tribes, religious minorities etc. in remote locations). Where budgets are insufficient, approaches combining durable and robust primary structure with temporary walling and cladding are appropriate to increase the cost effectiveness and reach of projects for the most vulnerable. However, decisions about the type, value and quality of shelter assistance cannot be taken in isolation from the capacity of beneficiaries to effectively use, maintain and upgrade their houses. Little support was offered to partners and beneficiaries to do so. Greater long-term improvements in safety and strength of buildings and greater support to partners and beneficiaries could have been delivered with more technical programme staff.
There were notable attempts, led by women, to deal with settlement-wide problems in an organised manner and to represent their largely disenfranchised communities to those in positions of power.
However, these were largely unsuccessful. There was insufficient attention as part the shelter projects, especially the relocation projects, to institute good governance and representation for communities. Had this been in place communities, and women, may have been more able so solve some of their lasting problems, even after projects end, funding disappears and NGO staff leave.
It is widely accepted that shelter projects will not be successful without addressing settlementwide issues. This study supports that, but furthermore highlights that from the point of view of most disaster-affected people in the locations studied it is livelihoods and WASH that most affect the wider success of projects:
• Shelter assistance delivered in combination with effective livelihoods assistance can have transformative effect, improving not only housing and incomes, but also education, health and other areas. In particular it can have an empowering effect on women and girls.
• Ensuring adequate access to safe water must be considered in shelter programmes. Several projects studied have resulted in communities without acceptable water supply, leading to poor sanitation and additional burden, especially on women and girls.
• Provision of toilets, without associated hygiene promotion programming, does not lead to changed behaviour or reduction in open defecation. Women and adolescent girls in particular suffer as a result.
Projects generally reflect the priorities of donors, government and NGOs and generally do not take sufficient account of the priorities of disaster-affected people. All the projects studied were agency-driven and largely contractor-built.
The form projects and shelters took was driven by donors, government and agencies and not by disaster-affected people. The funding available per household varies significantly and leads to great variation it the assistance delivered.
Robustness of buildings or speed of delivery has generally been prioritised over beneficiary choice and participation. Physical risk of future natural disasters has been successfully reduced, but other vulnerabilities have not been so well addressed.
There were examples of meaningful participation processes in which affected people felt able to significantly influence projects, leaving a lasting and positive impression. In most cases however communities are grateful for the significant support they have received, but do not remember being able to greatly influence the form it took. Largely due to the nature of post-disaster projects, but also due to insufficient consideration of how it could work, participation was less meaningful in the project design and beneficiary selection and more meaningful in the project implementation.
Donor mandates and priorities, coupled with insufficient technical understanding in agency programme teams, can lead to the almost arbitrary designation of houses as ‘temporary’, ‘permanent’ or even the highly confusing phrase ‘semi-permanent’.
These have the effect of obfuscating the true value and nature of what is being delivered. Temporary shelters are almost never temporary and no building is ever entirely permanent. Approaches that sought to maximise cost efficiency by designing buildings with durable primary structures and less durable cladding are entirely appropriate but were often lost in translation and not sufficiently understood, or agreed to, by beneficiaries. It is not appropriate to deliver ‘temporary’ buildings to vulnerable people without their understanding and without a viable plan to replace them.
None of the projects studies in detail involved specific consideration of the needs of disabled people, whether physically or mentally disabled. None of these projects involved specific consideration of the needs of elderly people. This has resulted in the needs of some of the more vulnerable people in society for safe shelter not being adequately met.