Chapter 6 Reconstruction Approaches

• Policy Makers
• Lead Disaster Agency
• Agencies Involved in Reconstruction
• Affected Communities

Guiding Principles for Reconstruction Approaches

• Households begin reconstruction the day of the disaster and government–guided by its reconstruction policy–may have to play catch-up in order that households and builders conform to, or participate in, any proposed reconstruction approach.
• Communities and households must have a strong voice in determining the post-disaster reconstruction approaches and a central role in the reconstruction process.
• The reconstruction policy must address the needs of households in all categories of tenancy: owners, tenants, and those without legal status. More than one reconstruction approach will probably be employed.
• The building approaches adopted after disasters should be as similar as possible to those used in normal times for similar households and should be based on their capacities and aspirations.
• Building codes and standards for reconstruction should reflect local housing culture, climatic conditions, affordability, and building and maintenance capacities, and improve housing safety.
• Reconstruction should contribute to economic recovery and the restoration of local livelihoods.
• Good planning principles and environmental practices should be incorporated, whatever the reconstruction approach.
   

For analytical purposes, this chapter makes a distinction among five reconstruction approaches that may be pursued after a disaster for the reconstruction of permanent houses. (A discussion of transitional shelter is found in Chapter 1, Early Recovery: The Context for Housing and Community Reconstruction.)

• Cash Approach: Unconditional financial assistance is given without technical support.
• Owner-Driven Reconstruction: Conditional financial assistance is given, accompanied by regulations and technical support aimed at ensuring that houses are built back better.
• Community-Driven Reconstruction: Financial and/or material assistance is channeled through community organizations that are actively involved in decision making and in managing reconstruction.
• Agency-Driven Reconstruction in-Situ: Refers to an approach in which a governmental or nongovernmental agency hires a construction company to replace damaged houses in their pre-disaster location.
• Agency-Driven Reconstruction in Relocated Site: Refers to an approach in which a governmental or nongovernmental agency hires a construction company to build new houses in a new site.

1. Government should decide on the policy for housing and community reconstruction, based on the results of the damage and loss assessment, and in consultation with the affected community and the lead disaster agency. Important decisions include: the reconstruction approach or approaches to be employed; the financial contributions to be made by various parties, including households; mechanisms for coordination; and the administrative and project management procedures that all agencies will follow.
2. Government should decide on the policy for reconstruction, based on the results of the damage and loss assessment, and in consultation with the affected community and the lead disaster agency. Important decisions include: the reconstruction approach or approaches to be employed; the financial contributions to be made by various parties, including households; mechanisms for coordination; and the administrative and project management procedures that all agencies will follow.  
3. Agencies involved in reconstruction should work with communities and government to agree on how the reconstruction policy will be applied, including the reconstruction approach or approaches to be employed and the financial contributions to be made by various parties, including households. These entities should also agree on general operational and project management procedures for reconstruction projects, without unduly restricting the ability of projects to be implemented.
4. The lead disaster agency should determine, in consultation with  government  financial officials, the level of assistance that will be provided for transitional sheltering, repairing, retrofitting, and reconstruction, and on the system for delivering funds. Government may want to impose a maximum assistance level for nongovernmental agency projects to reduce competition among agencies. See Chapter 15, Mobilizing Financial Resources and Other Reconstruction Assistance.
5. Agencies involved in reconstruction should agree with government on performance benchmarks for all reconstruction approaches and on reporting procedures, and collaborate on establishing the baseline and the monitoring system.
6. Affected communities should decide which reconstruction approach or approaches are most suitable for them and collaborate with government in the selection process. They should also decide how they prefer to organize themselves during reconstruction and should have the right to select which agencies will assist them and to agree on the form of assistance. Depending on the community’s political, social, and economic characteristics, organization of the community and collective decision making may require outside facilitation and support.
7. Whatever the approach, local governments must direct those aspects of reconstruction related to land use and physical planning and the regulation of construction. See Chapter 7, Land Use and Physical Planning.
   

Government should take an active role in setting the rules for and overseeing the activities of all agencies involved in reconstruction. It should provide the appropriate regulations and guidelines so that agencies conform to the following good planning and construction principles.

• Consistently apply good planning principles and conform to local development plans.
• Conform with local building codes and standards.
• Minimize environmental impacts in construction, site planning, and building design.
• Ensure community participation in all aspects of development, including those managed by outside agencies and private contractors.
• Maintain or improve the tenure status of households during the reconstruction process.
  

CA is appropriate for disasters that have a relatively limited impact and where housing damage was not caused by shortcomings in local construction practices. Emphasis with CA is on the distribution of financial assistance with minimal attention given to enabling measures. This approach may give affected people the choice to use the assistance based on their own priorities, which may not necessarily be housing. Some people may use the cash to migrate out of the disaster zone, for instance, if that is what they judge to be their best alternative.

However, the risks of ODR need to be understood and addressed. ODR requires good oversight and governance, that is, a government capable of establishing and enforcing standards, and some agency (governmental or nongovernmental) to ensure the quality of construction. Where engineered building technologies are being used, or multifamily housing is being rebuilt, using ODR is more challenging, but not impossible. The oversight from supporting agencies or government will need to be more technical, and experienced contractors must be hired. Success lies in establishing a support system for homeowners appropriate to the local context, which may include:

• Training of tradespeople and homeowners
• Technical assistance and construction supervision and inspection
• Updating and enforcement of building codes and construction guidelines
• Mechanisms to regulate prices and facilitate access to building materials
• A system for providing financial assistance in installments as construction progresses

Following the North Pakistan Earthquake of 2005, the Pakistani government promoted ODR to rebuild some 400,000 houses. Under the lead of the Earthquake Reconstruction and Rehabilitation Authority (ERRA), a multitude of international NGOs joined this program. Homeowners were responsible for the reconstruction of their own houses, with technical assistance and financial support disbursed in tranches. Insufficient capacity in the field can slow down the pace of construction and increase the likelihood of substandard construction work. To prevent this, ERRA facilitated the mobilization of decentralized teams who could provide technical updates and on-site training to the scattered beneficiaries. ERRA also used field observations and field testing to decide whether to allow different construction techniques and developed retrofitting methods to increase or maintain the seismic resistance of diverse housing styles. The approval of the local timber-frame construction style Dhajji was vital for the success of the reconstruction effort; statistical analysis indicates that, as compared to concrete block masonry, Dhajji houses are less costly and can be made acceptably seismic-resistant. Also, Dhajji construction techniques are easier for homeowners to understand, utilize, and adapt to local contexts, preferences, and resources. Three years after the earthquake, almost 300,000 seismic-resistant houses were nearing completion. An overarching factor in this success was the constructive way in which homeowners and those managing the implementation of the program were able to interact as the program was carried out.

CDR entails varying degrees of organized community involvement in the project cycle, generally complemented by the assistance of an agency. The degree of control over reconstruction by the community in CDR projects varies between agencies and from project to project. The agency may take the lead, suggesting housing designs, technologies, and/or materials, and delivering construction inputs and training. The agency may also employ skilled and unskilled laborers from the community or facilitate the formation of construction committees. At the other extreme, the community may manage most of the reconstruction process and receive only the support of facilitators (“collective ODR”). In summary, CDR may involve one or more of the following roles for the community:

• Organization and planning of the entire reconstruction process, including housing and infrastructure
• Decisions regarding housing design and building materials
• Production of building materials such as bricks
• Distribution of building materials or other forms of housing assistance (e.g., cash and vouchers)
• Hands-on reconstruction
• Oversight of builders

• Adopted by several national NGOs following the 2001 Gujarat, India, earthquake. The level of satisfaction was relatively high, but lower than for ODR houses.
• Used successfully as collective ODR following the 2006 Java earthquake in Indonesia. See the case study entitled Organizing Community-Based Resettlement and Reconstruction, in Chapter 12, Community Organizing and Participation.
• Adopted by the United Nations Centre for Human Settlements (UN-HABITAT), KfW, and Urban Poor Linkage Indonesia (UPLINK) in Aceh, Indonesia, following the 2004 Indian Ocean tsunami. Each of these agencies used a somewhat different interpretation of the approach. UPLINK gave people more choice in house designs, but community-based construction committees were given control over the purchase and distribution of building materials and over the mobilization of reconstruction labor. (In some cases, local contractors gained control of these committees.) KfW gave building materials and financial assistance directly to owners, but provided little choice over materials and designs.
• Used by the city of Ocotal, Nicaragua, to relocate and rehouse residents of displaced neighborhoods and highly vulnerable sites following Hurricane Mitch in 1998. Housing designs and building materials were proposed by a local architect, but receipt of a house was contingent on participation in construction of at least one family member. (See case study, below.)
  

After Hurricane Mitch struck the town of Ocotal, Nicaragua, damaging 1,164 houses and destroying 328, the mayor initiated a CDR project for resettling the affected population as well as for households located in high-risk areas. The guiding principle was to prevent future disasters by protecting the people, while improving the social cohesion of the community. The social dynamics of the community were carefully analyzed and community participation was promoted. The reconstruction process was explained to the citizens in community meetings, and the damage and loss assessment was conducted to reflect the community’s own priorities. Further, the new building site underwent an extensive planning process during which the proximity of the site to the future residents’ income sources was analyzed, as were possibilities for the future growth of the community, an important consideration when rapid population growth is expected.

Culturally and environmentally appropriate house designs, including improved traditional building materials and techniques, were proposed by a local architect and presented to the community. Future residents discussed the design and could request modifications, which were incorporated when technically feasible. Access to a house was contingent on full participation in the construction by at least one family member. Because Ocotal constructed its own adobe factory, it created much-needed employment in an effort to reduce out-migration from the town. Beneficiaries were trained in hazard-resistant construction, including the modification of traditional adobe building practices. Participation in the joint construction work on the building site made it possible for residents-to-be to establish initial contacts with their new neighbors. People’s pride and self-esteem increased as the project progressed, social cohesion was fostered, and a positive neighborhood identity was created. The Ocotal reconstruction project successfully incorporated prevention and built social capital, which has contributed to the sustainability of the project. In all, approximately 300 new homes have been built to date.

Because ADRIS takes place on the owners’ own land, it gives the homeowner some degree of control over quality, and sometimes the opportunity to participate in specific tasks. During construction, owners may be able to make suggestions to or modify the design. ADRIS eliminates the hurdle of land acquisition and generally allows the household to know where its house will be located. However, if housing designs are standardized or different from local designs, it may be difficult to fit the houses into pre-disaster settlement layouts or to modify them later. ADRIS, therefore, often results in similar or even worse outcomes than those of ADRRS, especially in the case of large-scale single-family reconstruction.

• Many international NGOs and private companies “adopted” villages and used ADRIS to build houses after the 2001 Gujarat, India, earthquake, even though government adopted an ODR policy. These projects often became a mix of ADRIS and ADRRS in adjacent sites where the housing designs did not fit existing sites and individual households, humanitarian agencies, or local governments bought additional land for new construction. In some cases, contractors did not respect the heritage sites and spatial organization, and caused irreversible damage to historical villages.
• Many private voluntary organizations adopted ADRIS in Tamil Nadu, India, following the 2004 Indian Ocean tsunami. However, they required that the land be cleared of houses and vegetation before starting construction. As a result, hundreds of pre-tsunami houses that were culturally and climatically appropriate and easily repairable were demolished, and thousands of trees were felled, which negatively affected people’s livelihoods and well-being.^[5]
  

However, for single-family homes, ADRRS can be problematic. It can lead to the construction of costly, inappropriate housing of poor quality and settlement arrangements that do not meet the socio-cultural and livelihood requirements of the people, causing severe economic consequences and low occupancy rates. The argument that ADRRS results in higher construction quality is rarely valid, because of poor supervision or the lack of qualified contractors. Moreover, finding an appropriate site can be a major challenge; failing to do so is, in fact, one of the principal reasons for dissatisfaction with this approach. The complexities of a decision to relocate are discussed in Chapter 5, To Relocate or Not to Relocate.

• International NGOs and national private companies opted for ADRRS after the 2001 Gujarat, India, earthquake because of perceived organizational advantages and higher visibility, including naming rights to new settlements. Local elites were sometimes given incentives to sell this approach to local officials. By accepting these offers, people lost their access to government financial assistance. When they later found the designs, layouts, and construction quality to be subpar and refused to occupy these villages, they ended up having to liquidate their assets, such as land and livestock, so they could rebuild elsewhere. An independent study found that in villages that opted for ODR, housing conditions were considered better than before the earthquake and economic conditions unchanged, while in villages reconstructed with the ADRRS approach, a significant percentage of households reported high levels of indebtedness and worse economic conditions.^_[6]
• ADRRS has had positive results in urban contexts. Two examples are the city of Nagapattinam in Tamil Nadu, India, and Banda Aceh, Indonesia, after the 2004 Indian Ocean tsunami. In Banda Aceh, a Korean voluntary organization acquired land in a middle-class neighborhood for an urban housing project. Although the houses were small, high occupant satisfaction was attributed to housing design, good location, access to public services, and the fact that livelihoods were not site-dependent.
  

When the Housing Foundation of the Islamic Revolution (HF)-United Nations Development Programme (UNDP) joint housing reconstruction project started following the 2003 Bam earthquake, the government of Iran and the HF (the executing agency for the reconstruction) had not fully defined the reconstruction approach. For the first year of the project, the HF hired contractors to build housing units for the program’s beneficiaries (129 female-headed households [FHHs]). But the poor performance and slow delivery by the contractors and their numerous claims for cost increases led the HF to shift after the first year to ODR with technical assistance.

The ODR approach followed several organized steps, namely, (1) submission of ownership documents or other verifiable proof of ownership in 1 of the 14 regional offices of the HF; (2) request for rubble removal from the property; (3) request for a demolition or leveling permit from the Bam Municipality; (4) delivery of a letter to the landowner by the HF office that introduced the landowner to the licensed consultancy firms that had established branches in the HF offices; (5) selection of a housing model from among those demonstrated by the private developers, contractors, UNDP, and international NGOs at the HF Technical and Engineering Site; (6) review and revision of the selected design with the consultancy firm until agreement on a final design; (7) receipt from the municipality of guidelines for engaging a contractor; (8) preparation of documentation for loans and grants from banks; (9) selection, negotiation, and contracting of a licensed contractor; and (10) commencement of construction. The beneficiaries received their first loan installment after the house foundation was complete. The shift to ODR resulted in more rapid reconstruction and higher satisfaction for the FHHs with the quality of the work.

Source: Victoria Kianpour, UNDP Iran, 2009, personal communication, http://www.undp.org.ir/.

Determining which reconstruction approach is preferable for an affected population—or even a subset of the population—is not a straightforward process. The disaster situation, and the conditions and preferences of households make each situation unique. This determination is also affected by the tenancy status of the household before the disaster and the desired tenancy status after reconstruction. However, some approaches may be more suitable to certain groups than others. The following table shows what may be the most suitable solutions for specific groups. It points out the importance of addressing the reconstruction requirements of owners who are landlords, since renters—a large proportion of the population in some countries, especially in urban areas—will be dependent on reconstruction by landlords. It is unlikely that a group of apartment dwellers (even if they were condominium or cooperative owners) would band together to reconstruct their units, particularly if reconstruction entailed relocation. However, this option is included here. More likely, they would liquidate their holdings and relocate elsewhere. The case study on the Gujarat earthquake, below, compares satisfaction levels of owner-occupiers with different reconstruction methods.

• Underestimating an affected community’s capacity to rebuild its houses and, hence, opting for reconstruction by contractors.
• Allowing those who can provide reconstruction funding to impose the reconstruction scheme.
• Building houses that people refuse to occupy for reasons of location, materials, design, or loss of livelihood.
• Not providing households participating in ODR projects with adequate assistance, facilitation, and supervision, resulting in poor construction quality, price inflation for materials, and other problems.
• Failing to take advantage of reconstruction as an opportunity to reduce risk and to strengthen local building practices and construction capacities.
• Inadequate oversight of private construction companies, which results in higher costs or inferior quality of construction.
• Designing and building houses that do not meet the communities’ cultural and individual requirements because of a lack of community participation in reconstruction planning.
• Local elites who hijack the project benefits because eligibility criteria and assistance schemes were poorly designed or not monitored during implementation.
• Pressure to overinvest in housing that leaves little or no funding for on-site investments such as infrastructure and restoration of natural habitat.
• Failing to provide sufficient technical assistance and facilitation to ensure that poorer households participating in ODR schemes reach construction milestones and obtain access to subsequent funding disbursements.
• In urban areas, adopting ODR without strengthening institutional capacity for land use planning, regulation, and building inspection, which can result in increased vulnerability.
• Neglecting the needs of tenant categories other than homeowners, e.g. owners of multiple family housing, tenants, landlords, and squatters.
  

1. When reconstruction is simple and mainly entails repair of damaged housing that is otherwise adequate, adopt CA; otherwise, whenever possible, adopt ODR.
2. Use CDR when community life and the local economy is disrupted by the disaster or relocation is required, or both.
3. Avoid ADRIS in rural areas and in places where the built environment and natural habitat are significantly intact.
4. If ADRRS is absolutely necessary, government should require community participation and establish simultaneous audit and oversight mechanisms.
5. Help communities rebuild their houses with facilitation and other appropriate enabling mechanisms identified through a social assessment that focuses on vulnerable households.
6. Ensure that reconstruction agencies take into consideration people’s different housing needs, vulnerabilities, livelihoods, and family size in selecting reconstruction approaches and that socioeconomic factors and gender-related requirements are addressed.
7. Under every approach, ensure that construction methods embody good planning, risk reduction, and environmental principles.
8. Require community participation in all aspects of the process, even when outside agencies or the private sector are in the lead.
   
   

In 2004, an independent household survey compared citizens’ satisfaction with different reconstruction approaches following the 2001 Gujarat, India, earthquake. The highest satisfaction was achieved with ODR with financial assistance and technical assistance from government, complemented by additional material assistance from local NGOs. All families whose houses were built using this model reported that their housing situation was better than before the earthquake. A second approach, government-supported ODR without NGO assistance, was almost as popular, with 93.3 percent of households reporting being fully satisfied. Relatively high levels of overall satisfaction (90.8 percent) were also reported under a third approach: local NGOs using CDR.

Contractor-built houses in Gujarat
Owner-built houses in Gujarat

When an earthquake struck the coffee-growing region of Colombia in 1999, national authorities worried about the repercussions of the disaster on the coffee exports-based regional economy. The President of Colombia created FOREC, a national fund that was put in charge of managing the overall reconstruction program. FOREC, in turn, decentralized the reconstruction process by distributing responsibility among 32 NGOs, putting each one in charge of a small town or a sector of an affected city. Rural reconstruction was assigned to the Coffee Growers’ Organizations (CGOs), a network of local, regional, and national committees represented internationally by the Coffee Growers’ Federation. However, the mission of a CGO was promoting coffee production and exports, not building houses or infrastructure. Lacking the means to implement a housing program, the CGOs opted for a user- or owner-driven approach in which beneficiaries were give responsibility for designing, planning, procuring, and building their own projects. FORECAFE, a rural reconstruction fund created by the CGOs, was charged with controlling the quality of construction on individual projects and managing progress payments, which were disbursed based on approval of the use of the prior payments. More than 14,000 individual housing, infrastructure, income-generation, and community services projects were completed in less than 18 months, thanks to an effective system of coordination of information, financial control, and quality management. This post-disaster, user-driven reconstruction experience (one of the first in Latin America) demonstrated the benefits of transferring responsibility over design, planning, and management of reconstruction directly to the individual beneficiaries of that reconstruction.

Source: G. Lizarralde, C. Johnson, and C. Davidson, eds., 2009, Rebuilding after Disasters: From Emergency to Sustainability (London: Taylor and Francis), http://www.preventionweb.net/english/professional/publications/v.php?id=11329.

In October 2005, a massive earthquake hit the Jammu and Kashmir region of India, killing more than 1,000 people and injuring 6,300. The impact on housing in some communities was catastrophic. In Tangdhar region, for example, 5,393 of 6,300 houses collapsed and 266 were partially damaged. In addition, winter was fast approaching, threatening to block access roads to the affected area. In contrast to many post-disaster situations where temporary shelters are a makeshift solution for a few months, sometimes built with inappropriate materials, the Jammu and Kashmir government decided to provide robust interim shelters. A reconstruction policy was needed that reflected local needs, priorities, and climatic conditions, including a proposal for the interim shelter construction approach. The Jammu and Kashmir government analyzed such options as (1) government construction of houses, (2) contracting NGOs to construct housing, and (3) facilitating construction by households, as was done in Bhuj, India, after an earthquake hit that city. The option chosen was ODR, and enabling mechanisms were established, including providing cash assistance of Rs 30,000 (US$677) for those whose houses had fully collapsed (enough for a 200 sq. ft. shelter) and sending engineers to survey villages and to help communities with technical issues. The transitional shelter design chosen could be built in two days. Although access to construction material was facilitated, people were encouraged to use lumber from their old houses to prevent shortages in the spring when permanent reconstruction work would begin. To ensure completion of shelter construction before winter hit, an incentive of Rs 5,000 (US$112) was given to the families that finished their sheds before the end of November while respecting safety norms. The reconstruction policy and technical advice were communicated to communities using flyers in Urdu and English with easy-to-understand drawings. In the end, 15,000 shelters—90 percent of the total—were completed by the end of November. A crisis was averted, thanks to a combination of a practical transitional shelter strategy, a clear message, good incentives, and strong support by the state for ODR.

Sources: Kutch Nav Nirman Abhiyan, 2005, An Owner Driven Interim Shelter Initiative in J & K. Report on Tangdhar Region, http://www.kutchabhiyan.org/PDF/InterimShelter_Initiative_in_J&K.pdf.

Adams, L., and P. Harvey. 2006. Cash for Shelter. Learning from Cash Responses to the Tsunami. Humanitarian Policy Group. London: Overseas Development Institute. http://www.odi.org.uk/projects/details.asp?id=367&title=cash-voucher-responses-tsunami.

Ayesan, Yasemin. 2008. External Evaluation of the Swiss Consortium’s Cash for Repair and Reconstruction Project in Sri Lanka, 2005–2008. Bern: Swiss Agency for Development Cooperation. http://www.sdc-cashprojects.ch/en/Home/Experiences/SDC_Cash_Transfer_Projects/Cash_for_Repair_and_Reconstruction_Sri_Lanka.

Barakat, Sultan. 2003. Housing Reconstruction after Conflict and Disaster. Humanitarian Policy Network Paper 43. London: Overseas Development Institute. http://www.odihpn.org/report.asp?id=2577.

Causton, A., and G. Saunders. 2006. “Responding to Shelter Needs in Post-Earthquake Pakistan: A Self-Help Approach.” Humanitarian Policy Network Humanitarian Exchange 34.  http://www.odihpn.org/report.asp?ID=2810.

Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ), 2003, Guidelines for Building Measures after Disasters and Conflicts (Eschborn: GTZ), http://www.gtz.de/de/dokumente/en-gtz-building-guidelines.pdf.

Duyne Barenstein, Jennifer. 2006. “Challenges and Risks in Post-Tsunami Housing Reconstruction in Tamil Nadu.” Humanitarian Policy Network Humanitarian Exchange 33. http://www.odihpn.org/report.asp?ID=2798.

Duyne Barenstein, Jennifer. 2006. Housing Reconstruction in Post-Earthquake Gujarat: A Comparative Analysis. Humanitarian Policy Network Paper 54. London: Overseas Development Institute. http://www.odihpn.org/report.asp?ID=2782.

Duyne Barenstein, Jennifer. 2008. “From Gujarat to Tamil Nadu: Owner-driven vs. Contractor-driven Housing Reconstruction in India.” http://www.resorgs.org.nz/irec2008/Papers/Duyne.pdf.