The international humanitarian community faced an unprecedented challenge in responding to the shelter needs created by the 8 October earthquake. The scale of the task was huge, with half a million homes in northern Pakistan damaged or destroyed. With winter approaching, aid had to be delivered quickly, and in difficult, mountainous conditions. Meeting these challenges required creative solutions and, especially, a high degree of reliance on local knowledge and ingenuity. Drawing on initial field evaluations of Catholic Relief Services (CRS)s self-help shelter programme, this article explores one approach to maximising local ingenuity to meet priority shelter needs. The approach was based on the understanding that households could find appropriate, efficient and durable solutions to rebuilding their homes.
Needs and challenges
The first challenge was the pace at which shelter needed to be delivered. The disaster struck only two months before the onset of winter in the northern villages. Located at the foothills of the Himalayas, higher-elevation villages can be cut off for days by several feet of snow. The previous year, CRS had responded to a winter emergency in these areas, providing food and blankets to remote villages that had been snowbound for over three weeks. In addition, the loosening of soil and rock on the mountainside after the earthquake raised the risk of further landslides.
The scale of the disaster was also unprecedented: 3.5 million Pakistanis were left homeless, exposed to freezing temperatures and rain. Moving to tented camps in accordance with initial government policy meant losing the October harvest, abandoning animals, leaving assets buried in the rubble and even losing land. Solutions were required in peoples communities of origin, especially at altitudes where tents would not provide adequate shelter against the winter weather.
Finally, the terrain presented significant challenges to the delivery of relief. The most vulnerable people lived in dispersed villages often accessible only by foot on narrow trails up steep mountainsides. Road access was possible along the valleys and to some higher locations with smaller 4×4 vehicles. Large-scale transportation of bulky shelter materials was problematic, and access was further disrupted by localised landslides. Attempting to coordinate distributions to scattered and difficult-to-reach villages required significant investment in logistical resources. In some cases, helicopters seemed the only option creative solutions were required to effectively reach higher elevations, solutions that could only be found by individual families themselves.
The shelter programme
The shelter approach developed by CRS was designed to maximise local ingenuity in overcoming some of the above challenges. A minimum set of material, financial and technical inputs were combined with social animation and mobilisation to enable families to build their own safe, adequate and durable shelters. The programme provided assistance to 20,000 households in North-West Frontier Province and Azad Jammu & Kashmir.
Material assistance was packaged as individual household shelter kits, comprising simple, lightweight materials, including roof sheeting, insulation, wire mesh reinforcement, fixings and tool kits, stoves and bedding. Families were expected to reuse timber from destroyed houses, and to salvage doors, frames and other materials for cladding. Instead of delivering shelter kits to each village, the CRS programme identified strategically located distribution points on routes into the target valleys. These points could be accessed from a number of different villages. Each household was then responsible for collecting and transporting materials back to their homes.
Since structured construction training was not feasible in the short time available, technical assistance focused on sharing simple messages on the safety, adequacy and durability of the shelters. In a central location within the village, and with the various households in attendance, a demonstration shelter was constructed by local carpenters to illustrate the key principles in practice. The construction time per shelter averaged approximately 13 days, and on average four people were involved per dwelling.
As a complement to material assistance, each household received a cash grant of 2,000 rupees ($35). No specific conditions were attached to the cash, though recipients were given general guidance on the range of intended uses. It was made clear that a portion of the cash was meant to enable recipient households to find their own means of transporting the shelter kit to their house site. Another portion was to support additional labour costs in building the shelter.
Findings of a field evaluation
The interim evaluation indicates that the programmes reliance on the resourcefulness of households, combined with material, technical and financial inputs, led to a range of initiatives for the transport of local materials, and a high degree of beneficiary and community solidarity extending beyond what a emergency shelter programme generally envisages. Providing cash for transport encouraged households to jointly hire trucks, use mules, organise family members in convoys and make links with villages in lower areas for the provision of temporary storage spaces.
Findings reveal that, on average, Rps1,900 ($33) of the Rps2,000 cash grant was actually used on shelter. Over 85% of households utilised the full amount exclusively on shelter construction. Of the total grant, the actual amount spent on transporting the shelter kits to the villages varied by geographic location. In one valley (Siran), households spent an average of Rps650 ($11) on transport; in another (Kontch), they spent Rps1,500 ($26). In some cases, local transport providers increased their prices in response to increased demand. Although this kind of market behaviour is generally considered one of the potential pitfalls of using a cash or market-based system, recipient households regarded it as an acceptable part of doing business.
The evaluation also revealed that households complemented the cash grant with considerable investment of their own. It was found that, on average, households spent an additional Rps6,600 ($115) on the construction of their shelter. As part of the shelter process, households were required to obtain their own timber, with an emphasis on salvaging timber from the debris of their original homes. Most of the extra spending went on additional timber, and sawing salvaged large-section timber beams into usable sizes at local mills. Sources of this cash included savings, loans, the sale of assets and compensation from the Pakistan government (although over 90% of the households had not received compensation before beginning shelter construction).
Cutting the beams into sections was important to ensure that the new roof was lightweight, as per the safety principles of the shelter approach. However, there was a shortage of sawmills in the villages. In the village of Akhori, families jointly bought a benchsaw and managed the cutting communally. In teams of eight, families transported the beams on their shoulders from the houses to the saw. Once construction was complete, the saw was to be donated to the local school, for use as a potential income-generating source, as well as safeguarding it as a village asset.
It was noticeable that, as the shelter programme was initiated in one village, households in adjacent villages immediately began to salvage timber and clear house sites, suggesting that the households recognised that the material, technical and financial assistance that was going to be provided would enable them to construct a viable dwelling. Many householders also commented that the materials, tools and cash were, with some additional investment, sufficient to enable people to build a more permanent home. This was clearly indicated in the decision taken by many households to cut their existing large beams (one of the primary assets of many households) into more manageable sections for use in their new home.
With substantial additional investment by individual households themselves, the majority of rebuilt shelters were adequate and durable, and will be the core or first rooms of new, permanent homes. The extent to which the provision of cash and informed technical assistance contributed to peoples decision to invest in their homes, and contributed to the degree of permanence and adequacy achieved, is difficult to assess. Clearly, households had access to resources that they were prepared to utilise, and basic building skills were available to undertake rudimentary construction. The injection of cash and a raised awareness of how to build appropriately expanded the range of options available to families to meet their shelter needs.
The contribution from beneficiaries themselves, in terms of materials and labour, was four to five times greater than the value of the package provided as part of the CRS shelter programme. Households themselves chose to develop what was initially envisaged as an emergency shelter programme into the beginnings of permanent housing, and also took responsibility for site planning, the sale or transfer of assets and the use of salvageable resources. In the immediate aftermath of the earthquake, given the constraints of pace, space and scale, it was felt that a highly complex logistical operation would be required, and affected households were assumed to have limited opportunities or resources to address their shelter needs. The CRS self-help shelter programme has shown that trusting in the resourcefulness and ingenuity of affected households themselves can prompt creative solutions to perceived challenges, and result in far greater long-term impact.
CRS Pakistan, Interim Survey of Self-Help Shelter Program, Siran & Kontch Valleys, February 2006.
Paul Harvey, Cash and Vouchers in Emergencies, HPG Discussion Paper, February 2005.
Graham Saunders, Housing, Lives and Livelihoods: Lessons in Post-disaster Assistance from Goma, CRS, February 2003.
Humanitarian Charter and Minimum Standards in Disaster Response, Sphere Project, 2004.
Alexandra Causton (firstname.lastname@example.org) is Head of Programming for CRS Pakistan. Graham Saunders(email@example.com) is Shelter and Settlement Technical Advisor within the CRS Emergency Response Team.