Pakistan, 2010, Floods

A.22 Overview Natural disaster

A.22 Pakistan - 2010 - Floods - Overview Overview Summary The 2010 monsoon season caused the worst flooding in Pakistan’s history, and one of the larger humanitarian crises of this century. The floods affected every province, over half of the districts in Pakistan, and one-tenth of Pakistan’s population. They damaged or destroyed 1.8 million homes, from the mountainous north where winters are cold, to the south where flooding caused the most damage. The scale was vast, but the funds did not meet the needs. For the first months, the government of Pakistan and many organisations working in the affected areas distributed tarpaulins, tents and other non-food items. The government also made cash payments to registered flood affectees using a “WATAN Card”. Following the emergency response, a “one room shelter” approach was adopted, by which organisations supported families to build a permanent shelter, which families could later extend. However the scale of the floods was such, that less than 10% of those who lost a house received such a shelter.

Before the floods inundated. Before the end of July, flood-affected individuals stayed Pakistan has a strong and over half a million people had with host families, 13% in col- recent experience of dealing with been affected and the emergency lective centres, 19% in planned humanitarian emergencies, from response began. camps, 10% in spontaneous set- tlements and 40% returned to, or conflict displacements (including It was another six weeks before remained in, their place of origin the Afghan refugee crisis and the the full extent of the floods became by September 2010. Many schools 2009 IDP crisis) to natural disasters known. The initial United Nations were used as collective centres. (with major earthquakes in 2005 floods appeal was launched as and 2008 and floods in 2007). As a waters were still rising in Punjab and Of the flood-affected areas in result there was significant experi- Sindh in the south of the country. Pakistan, Sindh province was the ence in dealing with the aftermath worst affected, with more than By mid-September 2010, the of disasters. However much of this 80% of affected houses either National Disaster Management was focused in the north of the heavily damaged or completely Authority (NDMA) estimated that country. destroyed, while in Punjab province the floods of 2010 had damaged 65% of affected houses were There were also significant stock- or destroyed 1.8 million households heavily damaged or completely piles of relief items some of which in Pakistan and that approximately destroyed. got flooded. Additionally there was 75% of the flood devastation was a manufacturing industry, being concentrated in Punjab and Sindh Response capacity one of the world largest manufac- provinces. turers of humanitarian tents and The disaster management ca- other key relief items. The floods led to wide scale pability of each affected province displacement. Some people were was quite different. In Khyber Pa- After the floods displaced for days. In other areas khtunkhwa province, where earth- The floods began in the north flood waters took six months or quakes and other natural disasters of Pakistan in late July 2010. longer to recede. occur more frequently, the Provin- Heavy rains lead to flash flooding, cial Disaster Management Authority landslides and areas becoming Surveys indicated that 9% of (PDMA) was relatively well-pre-

The floods caused significant damage to infrastructure and made 11 million people homeless. Photo: Joseph Ashmore

70 Over 1 million tarpaulins 380,000 tents were distributed. Additionally there was also a very large scale, though less coordinated, response from Pakistani civil society. Photo: Joseph Ashmore pared to coordinate and manage Recovery shelter In total less than 150,000 one the flood response. However, in The focus of the recovery room shelters and transitional Punjab and Sindh provinces which strategy was on the construction shelters were built. Although this is were the most affected, the ca- of one room shelters for those able one of the largest shelter responses pacities were much smaller because to return to their place of origin in history, it met only a small they had not previously managed a and transitional shelters for those fraction of the total needs. natural disaster of this scale. people who remained displaced, WATAN cards As temperatures in the north those with limited access to land, To support families during the were due to fall below zero a few and for seasonal migrants. relief and reconstruction phases, months after the floods, shelter One room shelter (ORS) was the Government of Pakistan estab- would become a lifesaving priority. defined as “a more durable solution lished a WATAN card scheme. The However, particular focus was built at place of origin with indig- WATAN Card is an ATM card that needed in the south as delayed enous materials and techniques.” the government was able to make flooding meant that the response The envisaged lifespan of the one payments to. It was distributed to was four to six week behind the room shelter was 3 to 5 years, which people in affected villages. north and there was an urgent can be extended upon upgrading of need to encourage an increase in the shelter. In the 12 months up to August capacity. 2011, the government of Pakistan Transitional shelter was defined issued WATAN Cards to 1.6 million Although strong national shelter as “a transitional solution that households, through which they coordination was established, the responds to temporary needs, e.g. received grants of 225 USD each. scale of the disaster was so large for those facing extended displace- that coordination efforts were chal- ment or those living in frequent A planned second phase would lenged, especially at district level. flooded areas”. Transitional shelters support 1.1 million households should have had a lifespan of at with cash grants of 450 USD per Emergency response least 1 year, and a design that household. The Emergency response was allowed for reuse of materials. relatively swift and on a very large scale, especially when taken in “The floods of 2010 are being termed as SUPER FLOODS comparison with other disasters. for Pakistan due to their large scale devastation in the country. Within the first six weeks of the They have laid new benchmarks for the country both in terms of response over 300,000 families preparedness and response. The floods affected 21% of cultivable had been supplied with emergency land of the country and uprooted 20 million people from their shelter items. homes and lands. Shelter, being the private asset, was once By the end of the emergency again the focus of attention for the government as well as the response, over 1 million households humanitarian community. It was important to help people get had been provided with a tent or back to their homes quickly to avert another food disaster in the two tarpaulins. country...” Despite the scale of this Waqas Hanif - Advisor National Disaster Management Authority (NDMA) response, it only amounted to 67% and focal person for Shelter Cluster of the total estimated need. These shortcomings were a result of the “Nobody was prepared for the magnitude. We were trying massive scale of the disaster, the to think big, but that was not enough. We went to provincial shortage of funds and shortage of coordination, to hub coordination, to district coordination; this experienced implementing partners had never happened before...” in the south of Pakistan. Arshad Rashid - Shelter cluster coordinator Pakistan floods www.shelterpakistan.org A.22 Overview Natural disaster PAKISTAN FLOODS IASCInter-Agency Standing Committee SHELTER CLUSTER PAKISTAN FLOODS RESPONSE AFTER 8 MONTHS BEFORE THE FLOOD Although the disasters compared in this document are very dierent in nature, human impacts, and challenges, this document makes some numeric comparisons based on data collected from the shelter cluster for dierent responses.

Area aected People who lost their houses Pakistan 2010 oods 11 million Pakistan oods 2010 100,000km2 DURING THE FLOOD - People are displaced to higher ground. 1.5 million Haiti 2010 earthquake The ‘slow tsunami’ that hit Pakistan in 2010 damaged or destroyed an estimated 1.7 million houses, leaving at least 11 million people homeless. In Punjab alone, twice as many houses were damaged destroyed by the floods than by the 2010 Haiti earthquake.

Spontaneous Collective Host families Camps 1,000,000 sites centres / rental only where Haiti 2010 <5,000Km2 750,000 absolutely

(drawn to scale) unjab Sindh necessary 500,000 P Houses KPK it Baltistan Haiti 2010 Note: Spontaneus sites includes dispersed shelter where one or two familes settle on elevated land near their houses. 250,000 K PA Balochistan Gilg

Camps require significant resources and can increase the challenges in return and recovery. Damaged / Destroyed Emergency shelter GAP map - 9th November 2010 0 RETURNS BEGIN Public information Advocacy Livelihoods suppor

EMERGENCY SHELTER RESPONSE

The cluster estimated that 70% of those with damaged and destroyed houses will need emergency shelter support. 381,000 tents 2.5 million delivered blankets Damaged Destroyed Spontaneous Host families delivered or on

Collective t

individual Ta Support for return Emergency shelter has been delivered to over 1 million their way houses houses sites centres / rental 1.36 million rgetting of vulnerable households. Despite the scale of response, only 67% of the emergency shelter needs have been met. tarpaulins delivered 1.5 million more required Tents not required for houses that can easily be repaired, nor for collective centres. s further gap lling still required for the Tool kits and Community clean up kits for damaged houses. remaining need Collective centres rehabilitated. Other Non-Food Items distributed are 438,600 bedding sets, Cash and vouchers should be considered where local markets are able to support the needs. 603,200 kitchen sets and 94,500 tool kits. estimated 30% of self recovery, with host-family support or in collective centre RETURNS CONTINUE - Some people have lost their land and are unable to return. Coordinatio Displacement tracking Damage assessment

SHELTER CLUSTER FUNDING 48% funded $168 million unmet requirement US$ 322 million requested n EARLY RECOVERY SHELTER RESPONSE s Damaged Destroyed Spontaneous Host families houses houses sites / rental Monitoring and review and needs Analysis of gender dif Longer term solutions Tarpaulins & rope, should be sought Construction materials Tool kits, Tool kits, Household kits Household kits cash Emergency: “One room” shelters As time progress families upgrade Community clean up kits Tents or tarpaulins Technical guidance one room shelters to form a home

As returns continue: ferences One Room Shelters are simple traditional structure made To date, Shelter Cluster members have constructed over 40,500 one room - Most people move to the land where their damaged or destroyed house was. from mud or brick, that allow families to upgrade and extend and transitional shelters and are committed to supporting the construc- - Collective centres and many spontaneous camps close. sites and buildings are rehabilitated. when they have the means to do so. tion of over 247,000 in total. - a rate similar to that following other major - Organisations consider transitional shelter as support. disasters. Transitional Shelters are lightweight structures that can Current commitments will, however, only meet 31% of the total need. be relocated. They are for those who cannot return to permanent land. Large numbers of families have started rebuilding on their own the Shelter Technical guidance for the ongoing self recovery process, Cluster’s priority is to support communities’ capacity for self-help. Other includes outreach messages and training of skilled labour and forms of support such as training centres are urgently required to local engineers. An illustrated version of the initial Shelter Strategy for Pakistan floods - 20th Aug 2010 provide technical assistance and help families to rebuild more safely.

Data sources: cluster websites for Haiti and Pakistan, FTS. The designations employed and the presentation of material on maps do not imply the expression of any 72 Some graphics reproduced courtesy of Stanford Kay Studio.com and opinion whatsoever on the part of the Secretariat of the United Nations concerning the legal status of JosephAshmore.org. any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. 31/03/2011 www.shelterpakistan.org Natural disasterPAKISTAN FLOODSShelter Projects 2010 A.22 IASCInter-Agency Standing Committee SHELTER CLUSTER PAKISTAN FLOODS RESPONSE AFTER 8 MONTHS Although the disasters compared in this document are very dierent in nature, human impacts, and challenges, this document makes some numeric comparisons based on data collected from the shelter cluster for dierent responses. Area aected People who lost their houses Pakistan 2010 oods 11 million Pakistan oods 2010 100,000km2 1.5 million Haiti 2010 earthquake The ‘slow tsunami’ that hit Pakistan in 2010 damaged or destroyed an estimated 1.7 million houses, leaving at least 11 million people homeless. In Punjab alone, twice as many houses were damaged destroyed by the floods than by the 2010 Haiti earthquake. Haiti 2010 1,000,000 <5,000Km2 750,000

(drawn to scale) unjab Sindh 500,000 P Houses KPK it Baltistan Haiti 2010 250,000 K PA Balochistan Gilg Damaged / Destroyed

Emergency shelter GAP map - 9th November 2010 0

EMERGENCY SHELTER RESPONSE The cluster estimated that 70% of those with damaged and destroyed houses will need emergency shelter support. 381,000 tents 2.5 million delivered blankets delivered or on Emergency shelter has been delivered to over 1 million their way households. Despite the scale of response, only 67% of the 1.36 million emergency shelter needs have been met. tarpaulins delivered 1.5 million more required further gap lling still required for the Other Non-Food Items distributed are 438,600 bedding sets, remaining need 603,200 kitchen sets and 94,500 tool kits. estimated 30% of self recovery, with host-family support or in collective centre

SHELTER CLUSTER FUNDING 48% funded $168 million unmet requirement US$ 322 million requested

EARLY RECOVERY SHELTER RESPONSE

Emergency: “One room” shelters As time progress families upgrade Tents or tarpaulins Technical guidance one room shelters to form a home

One Room Shelters are simple traditional structure made To date, Shelter Cluster members have constructed over 40,500 one room from mud or brick, that allow families to upgrade and extend and transitional shelters and are committed to supporting the construc- when they have the means to do so. tion of over 247,000 in total. - a rate similar to that following other major disasters. Transitional Shelters are lightweight structures that can Current commitments will, however, only meet 31% of the total need. be relocated. They are for those who cannot return to permanent land. Large numbers of families have started rebuilding on their own the Shelter Technical guidance for the ongoing self recovery process, Cluster’s priority is to support communities’ capacity for self-help. Other includes outreach messages and training of skilled labour and forms of support such as training centres are urgently required to local engineers. provide technical assistance and help families to rebuild more safely.

Data sources: cluster websites for Haiti and Pakistan, FTS. The designations employed and the presentation of material on maps do not imply the expression of any Some graphics reproduced courtesy of Stanford Kay Studio.com and www.ShelterCaseStudies.orgopinion whatsoever on the part of the Secretariat of the United Nations concerning the legal73 status of JosephAshmore.org. any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. 31/03/2011 A.23 Natural disaster

A.23 Pakistan - 2010 - Floods Case study: See A.22 “Pakistan - 2010 - Floods - Overview”, p70 for background.

Country: Project timeline Pakistan, Sindh. Disaster: 18 months - –– 21,700 shelters Floods constructed. Disaster date:

July 2010 No. of houses damaged: About 1.8 million in 77 of 139 districts across Pakistan No. of people affected: More than 20 million Project target population: Pilot project 175 households 8 months- –– Project completion Occupancy rate on handover: 100% as of 10 March 2011 Shelter size: 25m2 Materials Cost per shelter: 5 months - –– Project start/flood USD 740 Pakistan waters recede

2 months - –– Flooding reaches Sindh northern Sindh –– Relief operations start

July 2010 - –– Floods start

Project description This pilot project built 175 one room shelters for flood affected families in South Pakistan. It was later followed by a much larger scale project (building thousands of shelters over 18 months). Working through partners, the agency provided the construction materials and paid for skilled labour. Each shelter was built from burnt brick and had an accompanying kitchen and latrine.

Strengths and weaknesses 99 Conducting a pilot project allowed issues with the Although it was extended to 6 weeks, the actual time project process to be identified before a large scale taken to build the 175 shelters was 11 weeks. This was project was implemented. due to time taken to mobilise the community and rain 99 The returning families were supported to return to interrupting work. their original locations. 88 Lack of planning a delivery schedule and use of a 99 The community were consulted and involved single supplier led to the late supply of construction throughout the project, including the needs materials. assessments process, shelter design, materials 88 Lack of detailed specifications led to poor quality of sourcing and shelter construction. materials. This was compounded by lack of ownership 99 Skilled labourers from nearby villages also by the flood affectees leading to additional damage of benefitted from the opportunity to work on the materials during transport and unloading. shelter construction. 88 There was a problem accommodating labourers 99 The projects were flexible for the different needs from other villages, so tents had to be provided. of each village, depending on the social situation and 88 While the affected population were involved in the the preferred balance of shared facilities and privacy. project they could not focus on normal livelihoods 88 Once the project was started, the planned project activities, so the provision of food became an issue. duration of 4 weeks proved to be insufficient.

74 Natural disaster Shelter Projects 2010 A.23

The project was a pilot for a larger construction programme. Photo: Kpakpo

Before the floods original design: formal Non Objection Certificate See Background: A.22 “Pakistan (NOC) was obtained. In one of the • A high water table meant that - 2010 - Floods - Overview”, p70. three villages, all families owned the height of the foundation a piece of land, but in the other wall was increased from 150mm Selection of beneficiaries two villages the land belonged to to 450mm. A village assessment was carried a landlord, who provided the NOC. out to select three villages in one • 10% cement was added to district in Sindh for the pilot project. the mud mortar to increase its Implementation durability. The project was designed to A family assessment was then • At the request of the ensure a degree of participation by used to identify eligible benefi- beneficiaries, the pitch of the those receiving the shelters. Each ciaries. Each selected village had roof was reduced and the door family provided unskilled labour, between 40 and 80 families who dimensions were changed to and was responsible for plastering were eligible for the project. 1.2m x 2m. inside the shelter. The community Community mobilisation • As the sites were dense, the based organisations located and distances between shelters Community meetings were contracted skilled labour from were reduced and sometimes organised to discuss the needs of nearby villages. shelters joined. the returning families and participa- • To meet individual community In the pilot project beneficiaries tory rapid assessments were carried requirements, one village, were not paid. Their contribution to out. Community based organisa- consisting of one extended the project was to provide unskilled tions were established to manage family, built communal toilets labour and salvage materials. the projects. Members were trained and washing facilities. In other on mobilising their communities. Logistics villages where families wanted A single supplier was identi- The projects were coordinated more privacy, houses and fied following a tendering process and monitored through a district individual toilets were built in that included taking out advertise- coordinator and senior engineer, long rows. ments in the newspaper. However, regularly reporting to the District the contract did not stipulate the Coordination Office and the District Land allocation delivery schedule. As a result no Disaster Management Authority. In Before construction could materials were delivered in the first addition, each village had its own begin, field teams verified that 2 weeks of the project. site engineer and logistics assistant. there were no land disputes and a Shelter design The one room shelter was designed to the following brief: • minimum floor area of 25m2, • separate latrine and kitchen, • durable foundations, • brick/ concrete block construction with cement mortar. Due to site conditions and con- sultations with the beneficiaries, Shelters were built from brick with the aim of being more disaster resistant. modifications were made to the Photo: Kpakpo

www.ShelterCaseStudies.org 75 A.23 Natural disaster

During the procurement, material specifications only indicated the dimensions of the products and as a result, the quality of material varied. Furthermore there was lack of clarity over who was responsible for the materials once they had arrived on site. As a result, a lot of bricks and roofing materials were damaged during off-loading and moving. Families were expected to provide around 10% of the bricks that were required through salvaging materials. Design modifications Following the pilot project, it was agreed with the local authorities that future projects would include two structural improve- ments: • Walls should be strengthened by specifying cement mortar for the full height of the walls and not just the bottom 0.5m. • There should be greater resistance to earth tremors through the addition of a reinforced concrete ring beam at the top of the walls. The modifications increased the unit cost from around 740 USD to over 1,100 USD. In the 10 months after the pilot project, significant numbers (thousands) of one room shelters had been completed for the floods response. Materials list

Materials Quantity Excavation 8m3 Foundation: 4.6m3 3% cement & soil mixture Brick masonry (Plinth) 1:4 4.6m3 Damp proof course 0.52m3 Brick masonry (Wall) 1:4 16.13m3 Lintel (reinforced concrete) 1.7m (long) Wooden girder (roof) 12.3m Bamboo 83.4m Mat (2.4m x 6m) 3 Plastic sheet 6mx4m 1.5 20mm thick mud plastering 32m2 Door (wooden) 1 Woven mats 12 Different stages of the construction process. Houses were built with a bathroom and a kitchen. Photo: Kpakpo

76 Natural disaster Shelter Projects 2010 A.24

A.24 Pakistan - 2010 - Floods Case study: See “Pakistan - 2010 - Floods - Overview”, p. 70 for background.

Country: Project timeline Pakistan Disaster: 2010 Floods Disaster date: July – September 2010 16 months- –– 38,500 shelters No. of houses damaged: complete 1.8 million houses damaged or destroyed Project target population: 38,500 households 12.5 months - –– 20,000 shelters Estimated 217,617 beneficiaries completed targeted 11 months - –– 10,466 shelters Shelter size: completed Shelter sizes vary. 225 square feet (70m2) was recommended 10 months - –– 3000 shelters complete For mud structures, this was the suggested maximum Materials cost per shelter: 300 USD cash per shelter provided Pakistan If DRR recommendations are 5 months - –– Project start followed cost to beneficiaries 500 USD for a mud house, 4 months - –– Most people able to 1,000 USD for a fired brick return house July to September 2010 - –– Floods

Project description This large scale project provided cash to provide households with the means to buy materials and hire labour. Each household received the cash in 3 tranches. Each payment was made when a group of up to 25 households constructed to an agreed level. Payments were made via an agreed focal point for each group of households. The project was managed by 44 Implementing Partners spread over 3 provinces, most of them local agencies.

Strengths and weaknesses 99 Cash transfer allowed households to use money in 88 Because households were free to choose the the way they saw fit. construction materials they wanted, giving out disaster 99 Transferring cash instead of materials meant that risk reduction advice to each household was difficult. materials were purchased locally. 88 Not all of the implementing partners had the shelter 99 Using community focal points to distribute cash experience or the staff capacity to cope with the project proved to be overwhelmingly reliable. requirements. 99 25 households built shelters as a group, supporting 88 Some organisations working in nearby sites each other in order to receive the next payment. provided different amount of money, leading to initial 99 Disaster Risk Reduction trainings and messages to dissatisfaction among recipients and some drop outs. communities resulted in safer houses. 88 Internal requirements on financial accountability 99 A monitoring and evaluation and an information led to a significant amount of paperwork, requiring management system ensured that the programme 59,064 separate signed documents (various forms, was carefully tracked. MoUs, approvals, receipts, checklists, etc.). 88 The banking system in Pakistan lead to cash -- Identifying the most vulnerable households required transfers often being delayed. major efforts from the implementing partners and extra verification from the organisation.

www.ShelterCaseStudies.org 77 A.24 Natural disaster

Before the floods Beneficiary Selection Distribution in 3 tranches See Background: “Pakistan - 2010 The village committee was re- Each household received 100 - Floods - Overview”, p. 70. sponsible for selecting the most USD as an advance for digging Implementation overview vulnerable households using the the foundations and constructing following criteria: the shelter up to the plinth level. The project built One Room Once the implementing partner Shelters (ORS) through a cash 1) The house must have been had verified that all plinths had transfer system. One room shelters destroyed or heavily damaged as been completed, a request for the were traditional houses that could a result of the 2010 floods. 2nd tranche of 100 USD was made be extended later. They were 2) In addition, one or more of the and funds were subsequently dis- generally built with brick or mud following criteria were met: no tributed. walls. adult male in the family, disabled family member, medically unfit Once the Implementing Partner The project allocated a total family member, elderly family had verified that all walls had been of 300 USD to build each shelter. member, family taking care of completed, a request for the 3rd Households were then able to use orphans, large family, or poor tranche of 100 USD was made and the money to procure materials family. a final cash distribution was made and/or labour as they saw fit. The implementing partners for the construction of the roof. Technical advice was provided verified that the targeting had Cash transfer focal points to help families to improve their re- been done using the criteria before In each village, a representative silience to future disasters. people joined the programme. village committee was established. Monitoring staff also verified com- The 300 USD that was provided This was responsible for choosing pliance with targeting criteria. to each household was nationally a highly dependable and respected agreed between organisations in Groups of 25 households person from the community who November 2010. Some projects The organisation insisted that would act as the focal point for the that started later, or that worked in every household participating in project. areas with seismic risks provided a the project worked as a group and larger amount per shelter. The or- The focal point brought money completed each stage of construc- ganisation continued with 300 USD paid via his/her personal bank to tion before any household would per household because it allowed the village and distributed it to each receive the next tranche of funds. more people to benefit. group of 25 beneficiary households. The group of no more than 25 Many households added their When the focal point accepted beneficiary households had to work own resources to build their houses, the responsibility, he/she was given together as a unit. It was stressed in some cases selling assets such as a Memorandum of Understanding that none of the beneficiaries livestock. Many beneficiaries also to sign and told that bank charges would succeed unless all of them used emergency funds provided related to the programme activities succeeded. by the government through the would be covered. Upon the suc- “WATAN” card system. It was understood that they cessful conclusion of the project, had to help the most vulnerable he/she would also receive 175 USD. Most households participating in people to complete their shelter as the project built shelters that were It was explained that when ben- a precondition for getting money to significantly better than the house eficiaries signed the Memorandum construct their own shelter. they had been living in before the of Understanding to participate floods. in the programme, they were also

The project provided cash in three tranches so that people could purchase materials and hire labour. Photo: Usman Ghani, IOM Pakistan 78 Natural disaster Shelter Projects 2010 A.24 accepting the focal point to receive money on their behalf. Once the focal point had received the money, it was no longer the responsibility of the implementing organisations. The focal points brought three separate tranches totalling up to 2,500 USD each to the community. In only a handful of cases were funds not delivered to communities according to plan. This relatively small number of cases should be seen in the context that the programme worked with approximately 1,600 focal points. Housholder in front of her completed house. This was one of 11 photographs taken per household as part of the required project documentation. The 11 required photographs Photo: IOM Pakistan were: 1. Head of household standing Land rights is a major issue in In an effort promote transpar- in front of existing shelter Pakistan, especially in Punjab and ency and answer any questions 2. Head of household holding Sindh provinces which have large relating to the programme, posters CNIC (National Identity tracts of land under the control of in local languages were printed and Card) landlords. In this project, no dis- posted in the communities. These 3. Close up of CNIC card tinction was made between those posters had the phone number of 4. Head of household standing owning and those renting land. a call centre where people could in front of the empty plot obtain information and make com- Monitoring plaints. where they are going to Implementing partners were build required to fill in forms that es- Programme management 5. Head of household receiving tablished the vulnerability of the The programme operated in 3 first tranche (thumbprint on household, tracked the progress provinces. The programme head- documentation) of the construction and tracked quarters were in Islamabad but the 6. Plinth level completion the distribution of tranches to ben- day-to-day management responsi- 7. Head of household receiving eficiary heads of households. In bilities were devolved to four hub second tranche (thumbprint addition, the organisation had its offices. on documentation) own team that monitored around In Northern Pakistan, the or- 8. Wall level completion 7% of the households to verify 9. Head of household receiving ganisation directly implemented the targeting and to ensure that the construction of shelters, whereas in third tranche (thumbprint construction progress reported by on documentation) Punjab and Sindh provinces the or- the implementing partners was ganisation worked with 44 Imple- 10. Roof level completion being accurately described. 11. Household occupying the menting Partners, of which all but shelter A great deal of monitoring both 4 were local agencies. by implementing partners and the Technical aspects organisation ensured that house- After testing techniques during holds were meeting the construc- the pilot phase, multiple, highly tion thresholds. practical trainings for the im- GPS coordinates and 11 photo- plementing partners were held. graphs taken during the course of Different messages were developed cash distribution and construction for different types of construction. were required for each beneficiary Many field visits were made household so that construction to ensure that the messages were progress could be comprehensively being disseminated to communi- tracked throughout the process. ties and used in the construction. All of the information and pho- Three posters were produced that tographic evidence was uploaded showed the three main construc- into a large database that was Focal Points distributed money paid tion typologies and techniques that managed by a team of information via their personal bank account to could make shelters stronger and managers. each group of 25 households. more flood resistant. Photo: Carlo Gherardi

www.ShelterCaseStudies.org 79 A.25 Natural disaster

A.25 Pakistan - 2010 - Floods Case study: See “A.22 Pakistan - 2010 - Floods - Overview”, p.70 for background

Country: Project timeline Pakistan Disaster/conflict: Floods Disaster/ conflict date: July – September 2010 No. of houses damaged: 12 months- –– Project completion 1.8 million houses damaged or destroyed No. of people affected: More than 20 million people Project target population: 300 families Occupancy rate on handover: 7 months- –– Project start Unknown Materials cost per shelter: 316 USD The overall programme cost was significantly higher

Pakistan

August 2010 - –– Displacement in Sindh

Project description Provision of ‘One Room Core Shelter’ for flood affected vulnerable families in Jacobabad, Sindh Province, Pakistan. This project used a staged voucher system for beneficiaries to source all materials and to pay labour. This reduced logistical delays and greatly increased beneficiary participation. The design incorporated some disaster risk reduction considerations whilst still using predominantly local materials and practices.

Strengths and Weaknesses 99 Thick mud walls and roof keeps the inside of the 88 Material costs increased by 38% during the shelter cool during summer. implementation period. 99 Mud was sourced from the immediate vicinity, 88 Demands for skilled masons exceeded local supply. reducing logistical delays and environmental impacts. 88 Harvesting and planting seasons interrupted 99 Beneficiary led material procurement and quality construction. monitoring resulted in high quality materials. 88 Some local leaders tried to influence community 99 Vouchers redeemable for cash to make labour committees and suppliers for political reasons. payments allowed people to use the cash payments 88 Budget did not allow sanitation facilities to be built. for other requirements. 88 Variable security hindered monitoring activities. 99 This shelter design was acceptable to landowners. 88 Recruitment of qualified field staff was extremely The roof is the majority of the investment and difficult given the competition and scale of belonged to the families. reconstruction in the area. 99 The shelters were designed so that during flooding, 88 Structural integrity of the shelter highly dependent the mud walls could be washed away leaving the roof on good quality foundations. This was difficult to intact. monitor. 99 The shelter could be easily extended. Another 2 88 The project was unable to provide guaranteed columns and 1 girder would allow the shelter size to security of tenure for the recipients due to the immense be increased in size by 50%. power of local landlords and the entrenched feudal 88 Relatively small scale project given size of disaster. systems of landownership.

80 Natural disaster Shelter Projects 2010 A.25

The one room shelters were built using vouchers to pay for both materials and labour. Photo: Jake Zarins Before the flood Communities were prioritised the shelters were to be built. They See “A.22 Pakistan - 2010 - where the majority of buildings also led on shelter maintenance ini- Floods - Overview”, p.70. prior to the flooding had been con- tiatives, the planting of trees (flood structed using traditional materials break/plinth binding) and other Jacobabad district has tradition- (mud or ‘kacha’) and had been aspects of community safety and ally not been considered a high risk completely destroyed. improvement (e.g. danger of illegal area for flooding; it was only due electricity connections). to the exceptional flood levels in Additionally those villages with the 2010 floods (and again in 2011) higher than normal proportions Committees were also respon- that the area was submerged and of extremely vulnerable people sible for overseeing the selection population affected. and groups (single parent families, of vendors for the supply of bricks persons with disabilities, the elderly, and for the monitoring of delivered There were extremely high levels and those with no secure land materials. This group-led proof poverty before the flood with tenure or rights) were given priority. curement allowed communities both bonded labour and a lack of as a whole to reject poor quality land ownership for the majority. Selection of beneficiaries materials and negotiate timely This greatly increased the affected The implementing organisation delivery. This empowered commu- population’s vulnerability. used the following guidance on pri- nities and reduced the need for the After the flood oritising vulnerable groups for as- implementing agency to be present sistance: Jacobabad district was one of during each delivery and ensured the hardest hit areas during the • Poor families with three or more unscrupulous vendors could not flooding, with almost the entire children. take advantage. district submerged. An estimated • Women headed households. Beneficiary agreements 160,000 houses were destroyed • Households supporting orphans Prior to construction each ben- and as many as 1.4 million people or disabled families and eficiary signed an agreement clearly left without adequate shelter in chronically ill family members. stating the roles and responsibilities Jacobabad district alone. Under these criteria all benefi- of the beneficiary, the community, The bowl shaped topogra- ciaries selected for shelter assistance and the implementing organisation. phy prevented flood waters from were considered to be vulnerable. It highlighted that any deviation receding and much of the land Community committees from the prescribed process or remained submerged many months design would be solely their respon- Committees were formed con- after the initial flooding. sibility. sisting of beneficiary family repre- Selection of villages sentatives, village leaders and local The community committees Villages were selected through decision makers. were also responsible to ensure the conditions of the MoU were rein- close coordination with other hu- These committees, under forced and to assist project staff in manitarian organisations, govern- guidance from the community mo- dealing with any dissatisfaction or ment authorities and relevant local bilisation staff assisted in promoting complaints. actors to prevent duplication, and disaster risk reduction activities such also with the organisation’s projects as the raising of plinths on which in other sectors.

www.ShelterCaseStudies.org 81 A.25 Natural disaster

If the owners had to move they would be able to keep the materials in the roof. These were the majority of the cost of the structure. Photo: Jake Zarins

Technical solutions impossible to negotiate secure ciaries had undertaken work to The design was based upon tenure. the required quality, and reached brick and cement mortar columns. Despite negotiation, the land- the next stage in the construction Each column was built on an indi- owners refused construction of process. vidual foundation. These columns any form of full masonry (Pukka) Pre-selected vendors would only were aligned to support 3 steel structures. Under customary law receive payment once all vouchers girders and a traditional style flat the landlord automatically owns for a community had been signed roof of bamboo, reeds (Khick), any part of a structure sunk into the off by agency staff. Brick factories plastic sheet and mud plaster. ground. The roof which is the major engaged in the project were Beneficiaries chose whether to part of the value of the shelter monitored to ensure that no child build walls with either traditional would still belong to the families. labour was used. compacted mud or with sun dried The expensive roofing girders could mud bricks. in theory be removed and taken Implementation away if the family were ever evicted. Construction was completed in The shelter was finished with a comparatively short timeframe either a mud or a stabilised mud MoUs were agreed with land- once project preparations had been plaster on both the interior and owners prior to construction to finalised. The use of vouchers was exterior walls to provide protec- ensure beneficiary rights were extremely successful in ensuring tion from slow rising flood water or secured as much as practically timely delivery of good quality heavy rain. possible and would not be evicted materials. The communities policed from their homes to make way for In the event of serious flooding the process vigorously and did not other workers or families. and fast flowing waters the walls hesitate to reject any materials they between the columns would Vouchers considered to be of poor quality. dissolve leaving the roof intact. To increase participation, and in Any reloading and additional trans- consideration of a highly variable portation costs were the responsi- The approach used materials bility of the suppliers and they rarely and techniques that are familiar to security environment, a voucher system was used which also reduced attempted to supply poor materials the targeted communities. The thick more than once. mud walls help to keep the interior both the logistical burden and some of the Shelter cool even during the of the quality control responsibility At some stages in the project, extreme summer heat when tem- of the implementing agency. seasonal cultivation activities peratures rise above 500 C. Beneficiaries were trained in reduced the availability of labour. minimum quality requirements Land tenure During the 4 months of project of the materials and then issued implementation, inflation increased The majority of the families in with a booklet containing phased material prices by 38%. The use of the project were tenant farmers or vouchers for both materials pro- vouchers meant that the agency indentured labour who had for the curement and labour payments. could negotiate directly with most part occupied these areas for vendors for any adjustments in the generations. Due to the complex Vouchers could only be value of the redeemed vouchers feudal system of land ownership redeemed following sign off from without slowing construction. that dominates the region it was field staff who ensured benefi-