Table 1. Sixteen cases, reported in academic studies, are analyzed to draw analytical generalizations. Two main approaches and five possible scenarios are also identified.

Keivani and Werna argue for multiplicity in the provision of housing, that is to say, in the sharing of responsibilities and the organizational arrangements between the participants involved in housing initiatives. The participants might come from the public sector, the private formal sector, the private informal sector, or the residents and their community organizations. Previous research in our Group shows that procurement has a direct influence in the general performance of the building process (Mohsini, 1991, 1992, 1995; Abdel Meguid, 1997; Katsanis, 1998). In the specific case of post-disaster housing, we can also argue that both the organizational arrangements and the distribution of responsibilities have a direct impact on three important aspects in housing provision for the lowest-income families: (i) the institutional policy that lies behind the project; (ii) the level of local participation in decision-making; and (iii) the level of systematization of the building process. Figure 1 shows the general tendencies we found in reconstruction projects with regard to these three aspects.

Figure 1 also illustrates the two (plus one) ‘paradigmatic’ processes in which housing policies habitually fall (we have added, as number 1, spontaneous self-help, since it is adopted when other approaches prove to be insufficient):

1. Informal processes of spontaneous reconstruction. It is well known that permissive and flexible legislation (often actually indifferent towards housing needs of the poor) has resulted in the fact that fifty percent of the low-cost housing provision in the Third World is developed by the spontaneous initiatives of the residents. In the same way, evidence shows that survivors (particularly those in the lowest-income groups) usually begin their own reconstruction process shortly after the disaster (Davis, 1981). This case corresponds to scenario A, in blue.
2. Enabled processes of construction. Here we have used the term “enabled processes” to describe those initiatives that are developed in the framework of a reconstruction policy using a similar approach to the so-called ‘enabler’ housing policy (World Bank, 1994).
3. Provided finished processes. This one, obviously, refers to initiatives taken by authorities who directly assume the provision of accommodation.

Figure 1 also shows that, as was suggested in the first article, ‘enabler’ policies are frequently associated with a community-based approach, placing a strong emphasis on self-help and the participation of the informal sector. On the other hand, ‘provider’ policies frequently focus national and international initiatives on more technically sophisticated processes where the formal sector and specialized labor develop methodical processes of production.

It is important to clarify here that by ‘methodical processes’ we mean sequences of production activities that are deliberately correlated to attain a targeted objective (these activities are probably ‘systemic’ – but we prefer to reserve this word for approaches that blend various appropriate technologies, as we shall show). It is possible to state that in the spontaneous reconstruction housing process, both the participants and their actions are less coordinated and less able to attain the general objectives (at the scale of the community) than in the large-scale production of finished houses provided by international agencies.

Figure 1. Categorization of a group of reconstruction projects showing relationships between the organizational design of the reconstruction strategy and the different aspects of the building production process. The two main approaches are identified with green and gray. The possible scenarios presented in Table 1 are shown with letters (from A to E), and the different examples are represented by numbers, corresponding to the order in Table 1.

Observation of different cases also demonstrates that participation in local decision-making varies according to the organizational arrangement of the strategy used. It is important to highlight here that the general tendency of low-cost housing reconstruction strategies has been: “the more ‘provided’ and ‘methodical’ the process, the less local decision-making is permitted for affected residents”. This tendency is not surprising; in fact, Reinaldo Da Silva demonstrates that while in spontaneous self-help, the users make the majority of decisions (including designing the project and the administration of it), in organized self-help, the participation is almost exclusively restricted to the provision of labor (Da Silva, 1980).

However, we argue that it is possible to reverse this tendency in reconstruction strategies. In fact, we attempt to demonstrate here that an important contribution can be obtained with a strategy based on a pluralist approach. This approach aims to cumulate the positive results that can be obtained from previous strategies. In other words, we can suggest that this new approach might obtain the advantages of a community based approach at the same time as it obtains the advantages of a methodical process of industrialized building.

Decision-making

Let us deal first with the concept of decision-making in a pluralist approach. Obviously, decision-making needs to be accompanied by offering 'options of choice'. Indeed, in the successful after-disaster rural reconstruction in 1999 in Colombia, the NGO in charge emphasized a policy where survivors assumed the responsibility for choosing the most appropriate solution for each particular case from a ‘package’ of alternatives offered. These alternatives attempted to satisfy a great variety of needs and expectations. The evaluation of the program concluded that the positive results obtained were, in a great measure, due to this approach towards the community (Lizarralde, 2000).

It is well known in the post-disaster management literature that significant advantages can be obtained through the participation of the community in decision-making (Anderson, 1989; Davis, 1981). Although the same principle has, in theory, been transferred to post-disaster housing initiatives, its application has not always been sufficiently applied in practice. Evidence in the literature shows that rather than options of choice at different levels of the building process, it is choice at the level of the provision of labor which has been implemented - with more or less emphasis on self-help programs. Da Silva (1980), on the other hand, reminds us that self-help is much more than simply providing labor. According to Da Silva, there exist five potential types of self-help during the building process: (i) self-project preparation; (ii) self-project management; (iii) self-production of components; (iv) self-build; and (v) self-financing. Obviously, the participation in these five aspects can be combined to increase the involvement of the community.

The structure presented by Da Silva seems to be pertinent to discuss users’ participation in post-disaster reconstruction programs, particularly because frustration from reconstruction programs has come not only from the technical aspects of construction, but also at the level of architectural design, organizational design, logistics and, indeed, many other stages of the process (see the introduction of the article “Models of Reconstruction Projects”). If we follow the structure proposed by Da Silva, we can argue that a pluralist approach can be assumed at five different levels:

• Project preparation;
• Project management;
• Production of components;
• Construction (assembly);
• Financing.
  

Evidence exists to prove that great users’ participation in decision-making can be coupled with a systematic use of processes of industrialized construction. It is important to remember that the building industry in North America has, through the do-it-yourself move, paired an industrialized (and very methodical) process of manufacture of building components with strong users participation in decision-making.

The fact that post-disaster self-help initiatives are mostly targeted to basic upgrading and reconstruction whereas the do-it-yourself move is particularly associated with the renovations sector, usually at very different economic levels, makes it risky to compare both processes in terms of contribution to the building industry.

Self-help in developing countries has had a different consequence in terms of decision-making than the do-it-yourself move. If it is true that in North America industrialized products have facilitated users participation in building activities, in developing countries organized self-help has been frequently associated with a building process that is almost entirely done in-situ. This practice usually takes long periods for construction and makes use of small components made in an artisan way (some examples of this are the technologies based on adobe bricks, guadua or earth). In the cases in which a greater deal of ‘systematization’ has been introduced to organized self-help (in contrast to aided or informal self-help) the result has been the repetition of a standardized model that inexperienced labor force can copy without compromising the disaster-resistant standards. Obviously, this kind of standardization contributes very little to respond to particular needs of the residents.

We are aware that it is obviously inaccurate to see self-help exclusively as a way of involving the survivors in the reconstruction process. Defenders of self-help argue that training programs give residents skills to satisfy their accommodation needs in the future and skills to participate in jobs related with construction. Both of these aspects might be important contributions in the long run. However, if industrialized products combined with the do-it-yourself move has guaranteed that a great variety of “easy to assemble/install’ products are now available in the market for relatively unskilled labor in developed countries, in developing countries it has still to be proved that organized self-help can contribute to pre-occupancy flexibility and adaptability.

The approach to multiplicity requires the careful observation of particular needs in terms of financing, physical space, type of education and training, presentation of the information, materials available, technologies available, among others. We will not deal in detail with all of these variables. However, in order to illustrate the main argument of this paper, we will examine seven variables that can directly contribute to increase the range of choice and thus to reach more groups of survivors. The first five variables are related with project preparation and the last two with project management. These variables are:

1. Defining the outputs;
2. Acquisition;
3. Architectural flexibility;
4. Reaching the possible different scenarios;
5. Housing typologies;
6. Resources;
7. Information and education;

Defining the Outputs.

A post-disaster scenario requires building initiatives judiciously coordinated with mitigation actions at multiple levels. Observation of international cases shows that projects targeted to low-cost permanent housing must be articulated with other outputs such as:

• Emergency shelter solutions;
• Infrastructure repair (and possibly new provision);
• Community services repair (also including new provision);
• Repair of structures related with income-earning (small-scale commerce or industry, domestic commercial services, family business, infrastructure for agriculture, and so on);
• Employment opportunities for the community.
  

As a matter of example, post-Mitch hurricane reconstruction in Choluteca, Honduras, illustrates that the success of a housing project is dramatically altered when public services are not efficiently provided in coordination with shelter solutions, (see Anderson, 1999). In Choluteca, a relative success was achieved in shelter provision; however, according to Francois Audet, the local head of CECI (the development group that coordinated the projects) “[residents] may be living in the best house they’ve ever had, but we’ve got 5,000 to 6,000 people living in the desert without latrines, potable water or electricity”.

How to deal with immediate needs after the disaster while not affecting long-term development is a problem that has particularly exercised specialists in disaster management. However, according to the literature, very few imaginative solutions of outputs targeted to reduce the immediate loss after the disaster (providing protection and minimum comfort to affected families) avoid compromising the long-term sustainability of the project or, even worse, creating dependency towards external solutions. The successful core-housing project in the Philippines, carried out after Typhoon Sisang in 1987, is an inspiring example of an alternative solution to overcome this particular situation. In this case, a four units modular system that ‘grows’ according to the upgrading opportunities of the residents makes use of locally available materials. A training program supports the project in order to achieve a continuum building process, ranging from providing emergency protection to permanent housing. According to the resources of the family, this core-shelter is not only designed to be extended in area but also is conceived to be upgraded from indigenous materials to “more permanent and resistant” materials (Diacon, 1992).

Acquisition.

Housing policies in Latin American countries have been strongly criticized for concentrating almost exclusively on home-ownership, neglecting the potential of rental housing as a favorable solution for numerous lowest-income families (see Gilbert, 1997; Gough, 1998). According to Gilbert, “in Colombia, a rental housing policy is badly needed, given that 34 per cent of urban families rented homes in 1993 […] Certainly, no Latin American government has built housing for rent in the past 20 years […] most governments have failed even to mention rental housing in their policy statements”. Observation of reconstruction cases shows that the approach differs very little in post-disaster initiatives. In fact, (and this probably due to the strong emphasis placed towards self-help) UNCHS projects reported in Human Settlements and Natural Disasters (1989) failed to recognize rental housing in the objectives of the project-based programs, at least in a explicit way.

We argue that a pluralist approach requires the identification of different groups (‘targeted markets’) according to their expectations and their real-time possibilities of acquisition: Three groups might, then, be considered: (i) potential new home owners; (ii) potential new land owners; and (iii) potential renters.

Architectural flexibility.

Published in Harvard Design Magazine in 1999 Vikram Bhatt’s report of the Indian “Aranya housing project” developed in the eighties is an illustrating example of architectural flexibility in housing projects for low-income families (Bhatt, 1999). Bhatt associates the success of the program with the fact that special emphasis was made on the mix of different income groups, and providing different sizes of lots placed “on different sized streets to support mixed-use living and commercial activities.” This sites-and-services project addresses the diversity of ‘street’ activities. “More than just a transportation channel, the street is a place for work, shopping, and commercial activity, as well as the setting for social and religious ceremonies. Also several domestic activities […] occur outside individual house plots” (see figure 2).

Following in the steps of Bhatt (1999) we can suggest that in the study of reconstruction strategies, one should stress the importance of considering architectural multiplicity at four levels: (i) plot sizes (and even forms); (ii) different housing designs (in cases where designs are done or suggested by professionals); (iii) flexibility to adapt the house to different initial-occupancy needs; and (iv) flexibility in the house to grow in the future according to evolving needs or desires of residents.

Above: Figure 2. Aranya project in India. Model of typical sector. Note the diversity of housing solutions shown, the variety and quality of the public spaces and the general multiplicity of the architectural solution. (source: Bhatt, 1999)

Below: Figure 4. Study of typical mesones for the housing project in San Salvador (conducted by the Cooperative Housing Foundation, CHF). (source: Solo, 1991)

Above: Figure 3. General layout of the CIDA-CECI self-help based reconstruction project after Mitch (Honduras, 1999). Even relatively successful in its construction, this ‘rubber-stamped’ project is characterized by great monotony in the spread low-density layout, absence of community spaces and lack of variety and sensibility in the design.

Below: Figure 5. General view of the CIDA-CECI project in Honduras (1999). Source: <http://www.Colorado.EDU/hazards/wp/wp103/annexb/pic12.html>, March 19. 2001.

click to enlarge

Reaching the possible different scenarios.

Literature on post-disaster housing has particularly concentrated on projects of reconstruction of single-family detached housing. However, research shows that disasters affect different groups of residents with different needs and characteristics. Because this variety of affected population is scarcely discussed we suggest that it is important to develop a ‘systems analysis’ of the different possible scenarios of reconstruction. Furthermore, we will argue that reconstruction strategies must address particular solutions for each of the scenarios proposed. We mentioned, in the introduction, that there are four main groups to be catered for in post-disaster scenarios: 1. Non-affected housing. 2. Housing affected by the disaster but that represents safe conditions for the residents (repairing the house is more effective than reconstruction). 3. Housing affected by the disaster that cannot be occupied because it represents unsafe conditions for the residents (repairs are more difficult or more expensive than reconstruction; affected residents in this group might be homeowners or renters). 4. Pre-disaster deficit housing and homelessness.

The third group is, in fact, far from being a homogeneous one. Residents might be homeowners or not, the context can be a rural or an urban, the house might be demolished to build a new one or might be reconstructed using the affected one in part, etc. It is because of this variety of options that we suggest the analysis of four different cases, which represent different alternatives. Please click here: http://www.GRIF.UMontreal.ca/pages/a2body.html to see the different cases and sub-cases.

This variety of possibilities suggests that a complete reconstruction policy requires tackling the demand of both provisional and permanent housing for each of these cases – and possibly finding a way to tackle them within one systematic approach to the whole reconstruction task.

Housing typologies.

Very little evidence of the use of innovative and creative typologies in post-disaster housing is reported in the literature. The after-earthquake housing program in the Lio Village in Flores, Indonesia, carried out by the central Government, represents a remarkable failure due to the fact that traditional housing typologies were ignored. This case, reported by Gunawan Tjahjono (1999) in an article published in Environments by Design shows that not considering traditional typologies is a costly decision even in the lowest-income groups, particularly during the most difficult times of crises.

Ill-designed sites-and-services projects have been criticized for leading to rubber-stamp like projects where repetition, overcrowded houses and inappropriate distribution of private and public spaces are frequent (Bhatt, 1999). Self-help initiatives, that in response to a desire to avoid industrialized standardized designs attempt a better response to individual needs have, in many cases, not demonstrated the expected results. In many of these cases, unskilled labor is trained to develop a simple housing form that they can easily repeat without reducing disaster-resistant standards. During the time of construction, and in order to share the same interest in the construction of all the houses and to avoid particular preferences and disputes between participants, families usually do not know which house will be assigned to them. Houses are only assigned when the project is finished, and therefore it is difficult to believe that particular needs will be addressed before the occupancy of the houses. CIDA-CECI self-help reconstruction program after Mitch (1999) in Honduras exemplifies all these difficulties. Relatively successful in the building aspects, this monotonous and repetitive settlement of detached single-family houses can be characterized by its spread low-density layout, absence of community spaces and lack of sensibility in the design of the units, as mentioned above (see figure 3).

A cooperative housing project in San Salvador (conducted by the Cooperative Housing Foundation, CHF) shows the positive results that can be achieved with a sensitive approach to traditional typologies. CHF’s approach to traditional ‘mesones’ in downtown San Salvador enabled over 150 families to be housed in their own former meson sites by using an innovative approach of construction with the informal sector (see figure 4). By comparison, other approaches to the reconstruction of mesones, which were directed by municipal authorities, are reported by Tova-Maria Solo (1991) as less successful after introducing changes in the traditional layout of the units that were impossible to match with the economic viability of the project.

Housing literature is explicit in the advantages that can be obtained in terms of affordability and sustainability of low-cost housing when mid-rise typologies are used (see Shoenauer, 1994). However, this is an option that has been frequently underestimated in reconstruction projects. We consider that a re-evaluation of policies is needed to examine the potential contribution of mid-rise typologies in post-disaster programs.

Resources.

There seems to be a general consensus in the literature regarding the provision of different resources to the affected community. Indeed, evidence shows that reconstruction programs require addressing different needs of the survivors according to their existing (and very often insufficient) resources. Some of those resources might come in the form, for example, of materials, loans, subsidies, tax incentives, education and tents. This approach is perfectly illustrated by the remarkable initiatives taken during the post-earthquake Colombian reconstruction in 1999 to encourage the participation of large and well-established construction companies in the disaster area. In this ambitious program of rural housing, the Colombian authorities established numerous corporative and individual tax incentives targeted to promote employment opportunities in the affected area and to facilitate economic recovery. In fact, the organization responsible for the reconstruction program registered the creation of nearly 10 000 jobs one year after the disaster. Furthermore, with support of the private sector, an exhibition of seventeen model houses permitted the broad promotion of prefabricated buildings and prefabricated components, and demonstrated the interest of building companies in participating in the program.

Reconstruction resources might come from national or international sources. Despite the skepticism of some scholars towards imported technologies, technology transfer and international intervention, research demonstrates that international cooperation can be conducted in a way which supports local long-term development.

International partnership to deal with building initiatives is not new in the literature. W. E. Hewitt demonstrates in a recent article published in Habitat International (1998) that “increasingly, local governments in developing countries are entering into partnership arrangements with their counterparts in the developed world as part of an attempt to come to grips with problems associated with rapid urban expansion.” An illustrative example presented by Hewitt shows the effective results of the development assistance arrangements between the cities of Sao Paulo, Brazil and Toronto, Canada. In this example, information and technology sharing facilitated, at a very low cost, the transfer of a successful housing strategy in Toronto to contribute to alleviate the well-known housing crisis of the Brazilian city. Once again we argue that more effort is needed to involve pre-disaster research and international cooperation initiatives in post-disaster low-cost housing reconstruction.

Information and education.

As in the case of ‘resources’, there seems to be a general agreement among scholars towards a broad provision of information and education for the improvement of unsafe building practices. Ian Davis’ Disasters and the Small Dwelling (1981) (incidentally a publication where an extreme approach against industrialized imported solutions is adopted) presents convincing arguments concerning the importance and great responsibility of the media in a post-disaster scenario. Education and information are required not only for the affected community but also for the media themselves, potential donors and, of course, for the decision makers. In the context of a pluralist approach, where residents carry a great responsibility of choosing between possible solutions, an appropriate delivery of qualified information is the base of the reconstruction strategy.

1. Post-disaster low-cost housing provision is more than a technological problem. A successful reconstruction program targeted to low-income groups must not only consider the ‘hard’ factors of reconstruction (directly related to the building process) but also the ‘soft’ factors (not traditionally related with the building process, but crucial to attaining long-term development). Some of those ‘soft’ factors are: education, information, training, employment opportunities for the affected community, economic recovery, etc.

2. A reconstruction strategy must, ideally, be established before the disaster occurs. International exchange of information and research must be promoted and transferred to reconstruction policies in advance, in order to guarantee an appropriate response in times of crisis.

3. The success of a low cost housing reconstruction project is directly related to the provision of multiplicity and choice. In this way, the transfer of a great deal of responsibility for choice and decision-making to the affected community is related with the acceptability and the long-term sustainability of the project.

The keywords for the acceptability of a project are: choice, flexibility and adaptability and they must be considered in the architectural, technical, economical, and functional aspects of the project. Note that flexibility stands here for the short-term options (even before occupancy); whereas adaptability stands here for the long-term possibility to modify the dwelling after the occupancy of the project.

4. A reconstruction program requires (i) the coordination of national and international participants and resources and (ii) the articulation of ‘soft’ and ‘hard’ factors of reconstruction. Therefore, a careful organizational design is indispensable for the success of the program.

5. Only assessing the demand for physical accommodation hardly leads to solving the housing problem of developing nations. The solution requires a general policy that includes: (i) the provision and distribution of land; (ii) the provision of public services; and (iii) the provision of community services (schools, churches, retail, health centers, etc.).

6. Research demonstrates that some characteristics that appear spontaneously in informal settlements have not been seriously considered in housing initiatives targeted to the lowest income sector. Furthermore, research shows that incorporating these characteristics might positively influence the long-term sustainability of a post-disaster reconstruction project (which is our subject here), and therefore, their potentials should not be underestimated in low-income housing reconstruction strategies. We want particularly to refer here to four of these characteristics:

- The mix of families from different incomes;
- The mix of uses (particularly local scale commerce and services with housing use);
- The use of mid-density (3 to 4 story high) housing;
- The crucial importance of open and semi-open spaces in low-income settlements. The frequent use of these spaces (being private, semi-private or public) in domestic, social and commercial activities, particularly in tropical and subtropical regions.

7. A post-disaster reconstruction project requires the coordination of local and external resources, and both of these resources must be optimized to attain the long-term development of the affected community. However, if planned in advance before the disaster, international aid can, at the same time, be favorable to the particular interest of donor nations.

8. Rather than dismissing industrialized building because of certain extreme approaches which have often been adopted, we consider that industrialization, and particularly open systems of prefabricated components (if properly implemented) can positively contribute to achieve the sustainability and requirements of a post-disaster reconstruction scenario. A systemic view enables technology to be mobilized for appropriate short, medium and long-term community reconstruction, achieving a continuum of techniques that articulate local and external solutions.

9. Within the systemic view, post-disaster reconstruction can also be seen as a continuum from immediate relief to long-term rebuilding.

To summarize, we can say that while more research is needed to develop a coherent “international cooperation strategy”, there is already enough evidence to believe that a pluralist approach might overcome many of the difficulties found in current reconstruction approaches. Research is badly needed to broaden the current spectrum of approaches to post-disaster housing activities for developing countries, particularly in terms of project preparation, project management, production of components, construction and financing.

Abdel Meguid, Tarek Ahmed. Unpublished thesis. Université de Montréal. Managed claims procurement strategy (MCPS): A comparative study of the performance of alternate building procurement strategies. 1997.

Anderson, John. “An incomplete model town: New home for storm survivors in Honduras lacks basic services.” The Washington Post (December 28, 1999): A17.

Anderson, Mary and Peter Woodrow. Rising from the ashes: Development strategies in times of disaster. Paris: UNESCO, 1989.

Bhatt, Vikram. “Architecture for a developing countries.” Harvard Design Magazine (1999): 28-32.

Davis, Ian. Disasters and the small dwelling. Oxford: Pergamon Press, 1981.

Dian, Diacon. “Typhoon resistant housing in the Philippines: The core shelter project.” Disasters 16.3 (1992): 266-71.

Dudley, Eric. “Disaster Mitigation: Strong houses or strong institutions?.” Disasters 12.2 (1988): 111-121.

Gilbert, Alan. “On subsidies and home-ownership: Colombian housing policy during the 1990s.” Third World Planning Review 19.1 (1997): 51-67.

Gough, Katherine V. “House for Sale? The Self-help Housing Market in Pereira, Colombia.” Housing Studies 13 (1998): 149-60.

Habitat (United Nations Centre for Human Settlements). Human settlements and natural disasters. Nairobi: Habitat, 1989.

Hays, Alain. Construire pour la paix: des abris pour la guerre, des maisons pour la paix. Paris: Editions alternatives UNESCO, 1995.

Hewitt, W.E. “The role of international municipal cooperation in housing the developing world’s urban poor: The Toronto-Sao Paulo example.” Habitat International 22.4 (1998): 411-425.

Johnson, Cassisdy. Unpublished Thesis at McGill University. Temporary housing after Turkish’s 1999 earthquake. 2000.

Katsanis, Constantine J. Unpublished thesis Université de Montréal. An empirical examination of the relationships between strategy, structure and performance in building industry organizations. 1998.

Keivani Ramin and Edmundo Werna. “Refocusing the housing debate in developing countries from a pluralist perspective.” Habitat International. 25 (2001): 191-208.

Lizarralde, Gonzalo and Colin Davidson. “Models of Reconstruction Projects.” i-Rec: 2001 <http://www.GRIF.UMontreal.ca/pages/modelframe.html> Sept. 7, 2001.

Lizarralde, Gonzalo. Unpublished thesis at McGill University. Post-disaster low-cost housing: The case for social reconstruction. 2000.

Mohsini, R.A., and Colin Davidson. “Building procurement: Key to improved performance.” Building Research and Information. 19 (1991): 106-113.

Mohsini, R.A., and Colin Davidson. “Determinants of performance in the traditional building process.” Construction Management and Economics. 10 (1992): 343-359.

Mohsini, R.A., Sirpal, R., and Colin Davidson. “Procurement: A comparative analysis of reconstruction management and traditional building processes.” Building Research and Information. 23.5 (1995): 285-290.

Oliver-Smith, Anthony. “Post disaster housing reconstruction and social inequality: A challenge to policy and practice.” Disasters. 14.1 (1990): 7-19.

Ranaganath, Priya. “Mitigation and the consequences of international aid in post-disaster reconstruction.” University of Colorado, 2000. <www.colorado.edu/hazards/wp/wp103/wp103.html#tool> Feb. 18, 2001.

Roesch Da Silva, Reinaldo. Unpublished thesis Université de Montréal. La participation de l’usager dans le processus de construction. 1980.

Salazar, Alex. “Disasters, the World Bank and participation. Relocation housing after the 1993 earthquake in Maharashtra, India.” Third World Planning Review 21.1 (1999): 83-105.

Schoenauer, Norbert. Cities, suburbs, dwellings. Montreal: McGill University, 1994.

Solo, Tova-Maria. “Rebuilding the tenements: Issues in El Salvador’s earthquake reconstruction program.” Journal of the American Planning Association 57.3 (1991): 300-312.

UNDRO. Shelter after disaster. Geneva: Editorial of the Department of Humanitarian Affairs, 1993.

World Bank. Logement: permettre aux marchés fonctionner. Washington: The World Bank, 1994.

Yin, Robert. Case study research: Design and methods. London: Sage publications, 1984.