Collecting materials from the distribution centre Collecting materials from the distribution centre Photo credit: Victoria Maynard
Supporting shelter self-recovery: field experience following Typhoon Haiyan
by Victoria Maynard and Philip Barritt January 2015

Affordable housing experts have long argued that housing is ‘both the stock of dwelling units (a noun) and the process by which that stock is created and maintained (a verb)’.+J. Turner and R. Fichter, Freedom To Build: Dweller Control of the Housing Process (New York: Macmillan, 1972). They have also advocated for ‘supporting’ rather than ‘providing’ approaches: enabling families to upgrade their own housing situation through improved access to and management of land, finance, services, materials, skills and labour rather than the provision of completed houses.+N. Hamdi, Housing without Houses: Participation, Flexibility, Enablement (Colchester: ITDG Publishing, 1995). Over 30 years after Ian Davis described shelter after disaster as ‘a process, not as an object’,+I. Davis, Shelter after Disaster (Oxford: Oxford Polytechnic Press, 1978). process- oriented approaches to post-disaster housing are still rarely implemented on the ground.+J. Kennedy et al., ‘Post-tsunami Transitional Settlement and Shelter: Field Experience from Aceh and Sri Lanka’, Humanitarian Exchange, no. 37, 2007,

From policy to practice

This article documents the experiences of CARE International, Habitat for Humanity and their local partners in implementing self-recovery shelter programmes during the first six months after the Typhoon Haiyan.

Prior to the typhoon most affected families lived in light- weight timber buildings which they had either built them- selves, or paid local carpenters to construct. Many families had agreements with landowners which prevented them from building more permanent structures, while others were informal settlers in hazardous areas close to the sea. In response to Typhoon Haiyan, supporting half a million families to repair or rebuild their homes through a participatory process and the provision of construction materials, tools, cash and technical assistance was a strategic objective of humanitarian shelter agencies.+See Yet despite longstanding backing for support to self- recovery following disasters, and the appropriateness of this approach to the context in the Philippines, translating policy into practice was far from easy.

Following the distribution of emergency shelter kits (tarpaulins, fixings and tools), Habitat began providing construction materials to families in Cebu and Samar two weeks after the typhoon, including plywood, timber, corrugated iron sheets, nails and tools. CARE accelerated its planned shift from emergency support to support for self-recovery at the end of December as its assessments highlighted that many families had immediately begun to repair or rebuild their homes themselves. CARE’s self- recovery support included the distribution of construction materials, cash transfers through local cooperatives and household-level technical support. Both CARE and Habitat also offered community-level technical briefings on safe construction techniques.


Six months after the typhoon both agencies had assisted around 13,000 families to begin repairing or rebuilding their homes through their self-recovery programmes. According to the Shelter Cluster these were two of the largest such programmes implemented within the first six months of the response, with around 20% of households – 5% of the overall target for humanitarian shelter agencies for the first year of the response – provided with ‘support to self-recovery of shelter’ at that point. Programme monitoring completed by both organisations indicated that this assistance had enabled many families to quickly begin repairing their houses and move on to livelihood recovery.

Challenges and solutions



Existing partnerships meant that both organisations received rapid in-kind and financial support from major humanitarian donors in the immediate aftermath of the disaster. However, self-recovery programmes distributing construction materials had a higher cost per household than emergency shelter programmes distributing tarp- aulins. In the first few days of the response one donor required Habitat to reduce the cost of assistance provided per household (through reducing the amount of construction materials provided) in order to align its costs with those of agencies providing emergency shelter assistance. This imposed significant limitations on the impact of the assistance provided. In contrast, CARE was able to reallocate some funding from its emergency programming to self-recovery shelter programmes (although it was unable to reallocate funding already committed to emergency relief – see Figure 1). Some of CARE’s donors allowed the transfer of funds from emergency to self-recovery programmes, while additional funds were provided from CARE’s own fundraising efforts to meet the higher costs per household.

Beneficiary selection

In a context of limited resources and overwhelming need the process of identifying which families should receive assistance can be complex and time-consuming, and can cause tensions between families who are and are not selected. To overcome this challenge both organisations worked closely with communities to identify which households should receive assistance and provided relatively low levels of shelter support to almost all families in the first few months of the response. This approach ensured community involvement, supported the recovery of the whole community and reduced the likelihood of conflict. The simplicity of selection criteria and processes also ensured the minimum possible delay in the provision of shelter assistance, while more detailed assessments were undertaken to inform ‘top-up’ programmes later in the response.

‘No-build zone’ policy

In common with previous disaster response operations in the Philippines, the government quickly announced a ‘no-build zone’ policy prohibiting families from repairing or rebuilding houses within 40 metres of the coast. This caused confusion for families living within these areas (many of whom were informal settlers) and for organisations trying to assist them. One of Habitat’s donors would not permit the organisation to provide self-recovery support to families living within the no-build zones despite their urgent need for shelter assistance. This created significant conflict between families living inside and outside the no-build zones and eventually forced Habitat to work in a different municipality which had fewer families living within the no-build zones. More generally, Habitat was able to provide self-recovery assistance to families living within no-build zones after reaching agreements with each Local Government Unit (as they had done in previous disasters) and with support from more flexible donors. In general CARE chose to work in inland municipalities as initial assessments highlighted that these areas were underserved, but even in these areas some families were living in high-risk sites such as river banks, where they did not want to rebuild. In these instances CARE provided the household with shelter assistance and worked with the whole community to identify safe plots of land nearby.

Material procurement

Both organisations aimed to support economic recovery through local sourcing of construction materials, either through direct purchasing or the provision of cash grants. However, local purchase of construction materials in the immediate aftermath of the disaster was virtually impossible due to damage and disruption to supply chains. Habitat initially purchased construction materials from suppliers in Cebu, which was relatively unscathed, then identified local suppliers within the disaster- affected region as the construction industry recovered and expanded in response to demand. In Cebu demand from international agencies far exceeded supply in the first few months of the response. Suppliers were not able to fulfil orders on time or substituted inferior-quality materials, requiring Habitat to instigate higher levels of quality control. CARE purchased large volumes of items such as corrugated iron sheets in Cebu and Manila as these were not available in local markets in sufficient volume or quality to meet demand. CARE was able to rapidly procure due to early funding, initially avoiding problems, but experienced similar difficulties with later procurement rounds. Recognising that levels of damage, needs and capacities varied between households, CARE also provided conditional cash grants to be spent on house repair or reconstruction, allowing households to purchase materials or hire skilled labour, depending on their specific needs. This approach meant that families could prioritise how they spent the cash, and encouraged local processing of fallen coconut timber.+Over 33 million coconut trees were destroyed in the storm and processing the fallen timber before it started to rot and pest infestation affected undamaged trees was a crucial task. See


As with other agencies, both CARE and Habitat experienced transportation challenges in the first few weeks and months of the response. The transportation and distribution of construction materials was also more difficult than tents or tarpaulins due to their greater weight and volume. A combination of procurement delays on particular items, transportation challenges and a lack of secure warehousing meant that, in some instances, in the first three months after the typhoon Habitat chose to distribute partial rather than full packages of materials to affected families (returning later with the outstanding items). While this was a compassionate response to the needs of affected families in the immediate aftermath of the disaster it significantly increased the complexity of the programme and the number of distributions required. In later stages of the response both agencies distributed directly from trucks carrying complete packages to reduce the need for warehousing in the field. Despite CARE having made the decision to move from emergency to self-recovery support in the first few weeks of the response the organisation (and its local partners) was already committed to delivering its emergency shelter programme. This, combined with shipping delays caused by tropical storms in January, delayed significant scale-up of the organisation’s self-recovery programme until three months after the typhoon (see Figure 1).

Different needs, different solutions

Both CARE and Habitat initially provided the same package of assistance to all selected families. However, the success of self-recovery programmes is dependent on the level of damage and the capacity of each household to use the assistance provided. CARE’s partners brought valuable expertise in mobilising communities to help the most vulnerable, but nonetheless both organisations found that, in some cases, their support was not sufficient for families to achieve shelter recovery. In Guiuan, Habitat realised that many families needed additional support in order to continue with construction due to the extremely high levels of damage and lack of local capacity to recover. To meet this need Habitat worked with a small, flexible donor to develop a ‘top-up’ project providing a bespoke package of materials and labour to each household. This enabled families to continue rebuilding by combining salvaged materials from their pre-disaster home with additional materials donated by Habitat or other agencies, or purchased following the typhoon. Recognising that markets for construction materials were being re-established CARE provided a greater proportion of assistance in cash rather than materials when the organisation expanded its programme from Leyte into Panay. CARE also provided an additional ‘top-up’ cash grant and technical support to 4,300 particularly vulnerable households, with the support of the same donor from the earlier phase.

Safer construction

Academics, policymakers and practitioners advocate using the ‘window of opportunity’ created by a disaster to ‘build back safer’ and reduce vulnerability to future disasters. This is particularly difficult in self-recovery programmes, however, where the responsibility for building back safer lies not with the assisting organisation but with disaster-affected families – who may or may not view investment in increasing the resilience of their houses to future disasters as a priority compared to spending time or money reducing other risks. To overcome this challenge both organisations adopted participatory approaches to working with communities to encourage both understanding and adoption of safer construction techniques. Like other organisations, CARE and Habitat provided technical briefings to the whole community. In addition, CARE simplified technical guidance provided by the Shelter Cluster into a four-point catchphrase – ‘foundations, bracing, connections, roofing’ – to provide easy-to-remember minimum standards for both staff and communities, emphasised by printing on T-shirts and through other media including song and dance. To supplement the technical briefing CARE also paid ‘roving teams’ consisting of one community mobiliser and two carpenters from each community to provide ongoing technical advice for a two-month period, with frequent monitoring by CARE’s and its partners’ technical staff. Programme monitoring and evaluation indicates that CARE’s comprehensive strategy for the communication of safer construction messages resulted in high levels of knowledge and implementation of basic safer construction techniques, leading to safer buildings. It also found, however, that increased technical support and capacity-building of local partners could have further increased the robustness of the buildings and widened uptake beyond the direct beneficiaries of the kits.

From practice to theory

Every disaster response poses unique challenges. The experiences of CARE and Habitat following Typhoon Haiyan, however, suggest that a set of common factors within humanitarian organisations is key to the efficient and effective provision of support to shelter self-recovery, irrespective of the context.

Sectoral commitment and capacity

Habitat works solely on housing and settlements, while shelter is one of CARE’s four core emergency sectors, with a dedicated technical team to support its disaster response operations. Both organisations try wherever possible to support owner-driven approaches to post- disaster shelter. Institutional commitment meant that both organisations had high levels of technical capacity to rapidly deliver self-recovery programmes. This included standard humanitarian capabilities, such as financial management, quality control and logistics. It also included shelter and construction expertise at both strategic and operational levels, which enabled both organisations to identify appropriate responses and provide technical assistance to affected families.

Local knowledge

Habitat’s knowledge of pre- and post-disaster housing processes in the Philippines meant that the organisation rapidly identified an appropriate package of familiar construction materials. CARE’s existing partnerships with local NGOs and cooperatives enabled the organisation to adopt and implement a mixed material/cash-based approach at the speed and scale required. Both organisations were also experienced in working with communities, community- and faith-based organisations and local government to identify households eligible for assistance, manage the distribution process and negotiate local solutions to land and tenure issues.


CARE responded to the findings of its initial assessment by accelerating the switch from emergency to self-recovery assistance. Later in the response ongoing monitoring prompted both organisations to develop ‘top-up’ pro- grammes to respond to remaining needs. Critical to the ability of both CARE and Habitat to take this responsive approach was the support of their donors in reallocating emergency funding to self-recovery programmes (accepting higher up-front costs per household than emergency shelter programmes and recognising that savings would be made in avoiding on-going support or additional recovery funding), implementing ‘top-up’ programmes and providing assistance to families in no-build zones.

Based on the field experience of CARE and Habitat during the first six months following Typhoon Haiyan, it is clear that the barriers to greater adoption of approaches which support affected families’ own shelter recovery process are institutional, rather than conceptual. Tackling such barriers should be the focus of future research, policy and practice for implementing organisations, policymakers and donors.

Victoria Maynard is completing PhD research at University College London in partnership with Habitat for Humanity. Philip Barritt was CARE International’s Shelter and Reconstruction Advisor, seconded to CARE Philippines.