A study of housing reconstruction and social cohesion among conflict and tsunami affected communities in Sri Lanka

Purpose – The purpose of this paper is to critically analyse the role of housing reconstruction projects in post conflict and post tsunami Sri Lanka, and to discuss their implications on conflict prevention. Design/methodology/approach – Using four housing reconstruction projects in Batticaloa, Kilinochchi and Jaffna Districts, Sri Lanka, as case studies, and a novel methodological framework, the study explores the causal relations among the independent variables associated with housing reconstruction and dependent variables related to conflict prevention. The data, gathered from interviews and project reports, were analysed using propositions from a literature review, adopting a thematic analytical approach. Findings – This study finds that reconstruction has created new forms of conflicts and tensions for the people who came to live in the newly constructed houses. The hostile relations that existed among different ethnic groups during the conflict were continued, and to some extent, exacerbated by the reconstruction undertaken after the war. Practical implications – The study identifies causal relations among the independent variables associated with housing reconstruction and dependent variables related to conflict prevention, which can be used to inform physical reconstruction programmes after conflict. Disaster Prevention and


Introduction
A conflict that has been destabilising the social, political and cultural structures of a country over a protracted period of time does not end after formal hostilities have ceased. On the contrary, the hidden conflicts that lie beneath the political economic projection of the country often remain, and can come back to the surface if the root causes are not addressed through reconciliation. The post conflict reconstruction (PCR) process is a significant opportunity to seek reconciliation through promoting community participation, empowerment of people and addressing grievances.
Sri Lanka, attempting to recover after a protracted civil war arising out of ethnic tensions between the majority Sinhalese and the Tamil minority, was flooded with a large amount of aid and assistance, together with an opportunity to achieve recovery while addressing the root causes of conflict. It also opened prospects to spread development to the northern and eastern provinces of the country, which were largely neglected and also inaccessible during the war period. Some of these areas were severely damaged during the 2004 tsunami. Yet, due to the war and their relative inaccessibility, they received only limited assistance from the government and donors, when compared to other areas in Sri Lanka that were severely affected by the tsunami. A study by Kuhn (2009), commissioned by the Joint World Bank-UN Project on the Economics of Disaster Risk Reduction, found that expenditures in the housing sector were considerably higher, on a per-destroyed-home basis, in the Sinhala-majority Southern and Western Provinces than in the Northern and Eastern Provinces. Expenditures were particularly low in the most heavily affected areas of Eastern Province, Ampara and Batticaloa Districts. The study also found that some of this bias reflected the failure to convert funding allocations into expenditures, possibly due to the conflict, but it primarily reflected bias in initial donor allocations, with a bias towards the south and Hambantota in particular being consistent across bilateral, multilateral, and non-governmental donors. Bilateral donors tended to exert a strong bias against areas of Liberation Tigers of Tamil Eelam (LTTE) control while multilateral donors balanced these tendencies.
Consequently, after the cessation of the armed conflict, it was necessary to invest in reconstruction in order to compensate for the damages experienced and opportunities lost throughout the war years. Housing reconstruction was a significant priority among these reconstruction efforts, and received much attention from the local and international actors. However, the nature of housing reconstruction and the different stakeholders involved in the reconstruction activities created a dynamic socio political environment for the people who came to live in the newly constructed houses. This study is an analysis of the social, political and economic dynamics of the housing reconstruction in post conflict Sri Lanka, and its implications on conflict prevention. This paper is an account of a study to critically analyse the role of housing reconstruction projects in post conflict Sri Lanka and discuss their implications on conflict prevention. Section 2 describes the methodology adopted in collecting and analysing the data used for the present paper. Section 3 introduces the propositions used for the case study analysis through the literature review. The section 4 presents the findings of the study. Section 5 concludes the paper and provides recommendations.

567
Tsunami affected communities in Sri Lanka

Methodology
The data underpinning this study on housing reconstruction are part of a larger research study into the reconstruction of housing and infrastructure, and their effects on conflict prevention after the cessation of the armed conflict. The purpose of this wider project was to enhance the capacity of local stakeholders to deliver conflict sensitive infrastructure and housing reconstruction programmes within the North and East of Sri Lanka, and thereby help to prevent future conflict in the region. During the formative stages of the study, it was identified that a single reconstruction project goes through several stages: preparation, design, pre-construction, construction and use. These stages are based on the RIBA (2007) outline plan of work , which is widely used in industry, and helped guide data collection to ensure all aspects of the development process were considered by the research. The infrastructure projects were categorised into nine sectors, with housing identified separately, while recognising that some housing projects incorporate elements of infrastructure, such as roads, community facilities, and recreational areas. The nine infrastructure sectors were hard infrastructure (transportation, energy, water, communication, solid waste) and soft infrastructure (governance, economic, social, cultural). Within the post conflict environment, it was identified that these projects can act as either "dividers" or "connectors", as identified within the Do Know Harm Framework (Goddard, 2009).
A novel methodological framework was developed in order to guide the data collection. This framework is original in bringing together the characteristics of physical reconstruction projects and their potential to effect conflict prevention within the community, while also identifying sources of data and indicators/measures, both subjective and objective. Independent and dependent variables were identified through a series of workshops conducted with the participation of expert stakeholders from academia and relevant stakeholders, including governmental officials, donor agencies and representatives from the construction sector. The independent variables were identified with the project and related project components, while the dependent variables were identified related to the conflict, quality of life, economic and cultural impact. Using various data sources, subjective and objective measures were gathered for the pre-identified dependent variables. The methodological framework used for the overall study is given in Table I Table I.

Methodological framework
A case study approach was adopted in order to carry out the study. Several project sites were identified in the Batticaloa, Kilinochchi and Jaffna Districts of Sri Lanka, areas severely affected by the war and also home to many people who resettled after the tsunami and to different ethnic groups. The housing and infrastructure projects were chosen after several site visits with the participation of stakeholder officers. An exploratory case study approach was carried out in order to derive the causal relationships among independent and dependent variables, due to the lack of wellarticulated propositions from previous research. Randomly and purposely selected beneficiaries of the projects have participated in unstructured interviews conducted by project partners. Although the methodological framework provided a clear agenda and purpose for the discussion with the respondent, the unstructured interview helped to build rapport with respondents, getting respondents to open-up and express themselves in their own way, which was felt to be important due to the sensitivity of the context. It also meant that the research team was open to having its understanding of the area of enquiry open to revision by respondents.
The research design and instruments were all subject to the funder's and implementing institutions' governance and ethics procedures. In the design of this research, a number of ethical issues were identified as relevant to the project. The instability of conflict affected areas and the heightened vulnerability of populations caught in conflict, calls for careful consideration of the research methods employed, the levels of evidence sought, and ethical requirements.
Gaining informed consent was sensitive to the norms, customs and sensitivities of the local environment. Efforts were made to assess individuals in a particular group who were particularly vulnerable. A heightened level of sensitivity to a range of issues such as religious beliefs, cultural and social values, the legal environment and gender issues were considered. Wherever necessary, secondary sources were used as an alternative. The investigators were highly trained with a clear methodology and possibilities for referral for psychosocial support. Appropriate data management was employed to ensure a high degree of confidentiality with study data, in case such data could be used to support deliberate targeting of individuals/groups by perpetrators of violence. The study was also designed to ensure that the participating communities will benefit from the research by a focus on specific practical applications of the research findings.
Among the selected projects for the overall study, four housing projects were used as cases for the purpose of this housing project focussed study. The projects were selected to cover a range of districts in both the north and east of the country, with different ethnic mixes and donor arrangements. The data gathered from interviews and project reports were analysed using the propositions from the literature review and identified within the methodological framework. This paper uses four housing reconstruction projects as cases, adopting a multiple case, holistic approach, as evidence from multiple cases enabled a wide variety of project characteristics to be considered (as detailed in the methodological framework, see Table I). The unit of analysis is the housing reconstruction project. Propositions based on the literature can be used to analyse cases in relation to their political, social, historical and personal contexts, which are significantly contribute to understanding the cases in detail (Baxter and Jack, 2008). The analysis explores the causal relationship between housing reconstruction and conflict prevention. Linking data to propositions is done through explanation building, and the linkages between propositions and case study data are used to develop the analysis. The analysis method of linking data to propositions 569 Tsunami affected communities in Sri Lanka allows a focussed analysis, exploring rival propositions, providing alternative explanations, and increasing confidence in the findings (Yin, 2014).
A thematic analytical approach was followed, with data analysed according to the themes identified through the literature review. Thematic analysis is a method used for description and interpretation of data using both inductive and deductive arguments. It is a non-linear analytical method that emphasises heavily on the context and integrates the content. While being a method of identifying, analysing and reporting themes within data, this approach is purely qualitative and allows a detailed narrative of the data (Vaismoradi et al., 2013).

Literature review 3.1 Disasters
The United Nations defines disaster as an event that critically disturbs the functioning of a society. It can cause human, material and environmental damage, making it difficult for societies to resume back to normal conditions on their own (United Nations International Strategy for Disaster Reduction, 2009). In his typology of disasters, Robinson (2003) includes conflicts and wars within the category of complex emergencies, which are often caused by human actions and results in large scale mortality, displacement, human rights violations and food insecurity.
A common feature among all kinds of disasters is that they bring about large scale destruction. Disaster is typically followed by death, injury, disease, negative impacts on physical and mental well-being, damages to property and environment, destroyed assets, and economic downturns (United Nations International Strategy for Disaster Reduction, 2009). Disasters have a significant impact on poor and developing countries. They have long-term impacts on social and economic systems due to the destruction of infrastructure and displacement of populations. The vulnerabilities that already existed in the communities will determine the severity of impact from disasters (McDermott et al., 2014).

Reconstruction
Reconstruction is generally referred to as an innovative approach to solve issues that are crucial to development after a crisis (Brun and Lund, 2008). The process of post disaster reconstruction embeds strong decision making using limited information in order to make long-term impacts. It is necessary to coordinate the inflow of resources to achieve tangible goals within a limited time period (Guarnacci, 2012). Reconstruction is not merely reviving back to a pre-disaster condition, but also addressing the existing vulnerabilities and empowering the communities to face future disasters. This is widely known as "build back better" ( Jordan et al., 2015). The international community is increasingly interested in the sustainability of reconstruction efforts. Guarnacci (2012) defines sustainable reconstruction as a long-term strategy which considers sustainability under different but interrelated criteria. Although there is immense pressure on governments and external donors to rebuild quickly, reconstruction is a process that requires extensive planning and therefore a long period of time ( Jordan et al., 2015).

Reconstruction in a post conflict context
Conflict does not terminate immediately after the cessation of armed hostilities. On the contrary, they can often continue for a long period of time (Brown, 2005), which may leave underlying tensions and causes unresolved. The end of a war, especially by military intervention, does not guarantee peace. Gellman (2010), in his study of PCR in Cambodia, noted that after the official declaration of the end of the war, post conflict countries typically undergo a period of political instability, economic marginalisation, and hidden violence.
In a conceptual framework for PCR, Jabareen (2013) asserts that reconstruction offers a greater opportunity to reduce the risk of conflict. This mirrors the findings of studies in Malaysia, Northern Ireland, Mozambique, and Guatemala (Stewart, 2005), which found that policies need to be adopted with political sensitivity, and it is important to understand the roots of conflict before developing reconstruction policies.
Others highlight the broad impact of reconstruction. Drawing upon evidence from evaluation studies in Afghanistan, Bosnia-Herzegovina, Croatia, East Timor and Rwanda, Anand (2005) suggests that by coordinating among different policy interventions and stakeholders, reconstruction can bring successful solutions to developmental challenges faced by post conflict societies.
In Africa, Theron (2011) identifies reconstruction as a significant process on the road to recovery, and thereby to build sustainable peace in the presence of issues related to security, political governance, socio-economic development, gender and justice. Likewise, in a study of post conflict risks, Collier et al. (2008) identified that the two major challenges faced by post conflict societies are typically economic recovery and preventing future conflict.
The role and nature of reconstruction in Sri Lanka has been highlighted in a number of studies. Höglund and Orjuela (2011) caution that societies such as those found in Sri Lanka at the end of the armed hostilities, are in the danger of going back to conflict. They stress that conflict prevention must be a major focus in the post war economic and political agenda in Sri Lanka. Brun and Lund (2008) observed that the war situation changes the power relations of different communities and the capacity of vulnerable communities to make decisions for themselves, meaning that reconstruction in a conflict context requires careful consideration of the war dynamics and context. Prior to the end of the war but in the aftermath of the tsunami, Rajasingham-Senanayake (2005) described a process of recovery with a broader approach of involving communities to build their capabilities to deal with post crisis situation. The conflict context is different from the intervention in non-conflict contexts. If the reconstruction process is insensitive to social, political and cultural dimensions, the society may be drawn back to conflicts, sometimes after decades.
The term conflict prevention suggests different things to different people and there is no agreed-upon meaning among scholars. Much of the discourse has an emphasis on preventative diplomacy, but for the purpose of this study, the Carnegie Commission's (1997) definition will be used, which covers the conflict affected and the emerging conflict: "The aim of preventive action is to prevent the emergence of violent conflict, prevent ongoing conflicts from spreading and prevent the re-emergence of violence".

Housing reconstruction in a post conflict context
Housing is an essential and complex asset linked to livelihoods, health, education, security, and social and family stability. Barakat (2003) describes the house as a social centre and a form of identity. Similarly, Barakat and Zyck (2011) note its role as a unique form of dwelling that forms the social, economic and political realities of the individuals. Dale (2015) highlights housing as an indicator of ownership of resources, and being a property associated with legal ownership, housing differs from all other types of relief. The destruction of this highly vulnerable asset is significantly more visible that other types of infrastructure during conflict (Barakat, 2003).

571
Tsunami affected communities in Sri Lanka After a residential area is severely damaged due to war, reconstructing houses to meet pre-war conditions is difficult (Dale, 2015). While being an essential part of the recovery programme, housing reconstruction is linked to social and economic recovery of the community. It is also an indicator of the extent of recovery ( Jordan et al., 2015). Therefore, housing reconstruction forms a major part of post conflict and post disaster agenda.
Housing is often treated as a development concern, whereas it is also a humanitarian issue, particularly in the post conflict setting. Based on reviews of experiences in housing reconstruction in the aftermath of disasters caused by natural hazards and conflict, Barakat (2003) stresses that building houses is a process of complex planning and long periods of preparation. Barakat and Zyck (2011) observe that it requires a greater extent of contextualising that other physical infrastructure. Similarly, a study by French (2015) cautions that the manner in which the management of housing, land and property rights are ensured is a crucial determinant of sustaining peace in a post conflict society.
Houses are connecting the communities socially, rather than standing in isolation to one another. They can cause conflicts as much as they can resolve them (Brun and Lund, 2008). If one group is favoured in terms of housing, it can be an indicator of dominance of a certain group in society over another, particularly after a period of division and discrimination (Dale, 2015). Housing reconstruction is often politically sensitive, as it is a meeting point of geopolitics and identity politics (Brun and Lund, 2008). Therefore, the relationship between state and the citizens becomes extremely important during post conflict housing reconstruction (Dale, 2015). The present practice of housing reconstruction often recreates and even exasperates the existing vulnerabilities in terms of disaster, poverty and gender. Assistance provided for housing often benefits the wealthier or better off communities (Barakat and Zyck, 2011).
Reintegration and return of internally displaced persons (IDPs) is critical in achieving peace in post conflict societies. In order to facilitate and accommodate a large number of returnees, an equal amount of housing reconstruction is essential (Stefansson, 2006). Housing and property restitution is a significant part in the return and resettlement process (Smit, 2006). Respecting the legal and moral right of the displaced to return to their homes is an important step to rectify for the discrimination and ethnic cleansing they may have experienced throughout the war period (Stefansson, 2006). House and home making are interlinked concepts for marginalised groups. In order to support themselves with a livelihood, the displaced people should first have a house (Brun and Lund, 2008).
After the 2006 war in Lebanon, an owner-driven approach was adopted by several Middle Eastern donors through directly granting compensation to owners to build houses. The approach resulted in the building of a large number of houses and a high return rate (Barakat and Zyck, 2011). In certain countries, where the economy is market driven, it is typical to prioritise rebuilding industrial plants over housing reconstruction, as it was the case of Russia after Second World War. The period after the war was considered as an opportunity to build new cities and bring about reforms to social, economic and political systems (Dale, 2015).
However, it is important to note that some of these reconstruction efforts created several negative impacts. The majority of female headed households experienced severe delays in receiving compensation. Women, being specifically vulnerable, typically face severe difficulties during reconstruction process (Barakat and Zyck, 2011). In post conflict Russia reconstruction and resettlement were not properly planned in line with the main agenda regarding the growth of the city, thus making a long-term impact on development (Dale, 2015). The lack of technical assistance for houses constructed by owners in Lebanon, resulted in low-quality housing making them vulnerable to disasters caused by natural hazards (Barakat and Zyck, 2011).
In a general context, people are willing to return to their pre-war houses due to morality and justice issues. Yet, the meaning of house and home can significantly change due to the effect of war and displacement, as noted by Stefansson (2006) in a study of Bosnia and Herzegovina. IDPs tend not to return to their home, especially if they belong to ethnic minority groups. As a result, large scale forced resettlements were carried out as government initiatives. The property restitution pattern fuelled the ethnic separation in Bosnia and Herzegovina. Stefansson argues that although the international community viewed return of Bosniaks as a form of peace and reconciliation, it ultimately resulted in increased ethnic divisions and isolation of minorities.
These international examples emphasise that housing reconstruction after conflict has implications on the lives of people beyond the mere purpose of providing shelter. According to Brun and Lund (2009) the present discourse of literature largely discusses housing as a means of contributing to development and strongly emphasises the physical dimensions of housing. Housing interventions have the ability to improve the physical and emotional resilience of people, and facilitate reconciliation through participation and recognition. It should be given more consideration in a post conflict and post disaster agenda (Barakat, 2003).

Post conflict Sri Lanka
According to the United Nations definition, both tsunami and war can be categorised as disasters. Sri Lanka experienced a period where these disasters coincided. While suffering from protracted ethnic conflict from 1983 to 2009, Sri Lanka was hit by the Indian Ocean Tsunami in 2004. The tsunami damaged more than 95,000 houses (Nakazato and Murao, 2007). Moreover, by the end of the war in 2009, 280,000 people were internally displaced (UCPD, 2015). It is evident that housing was a sector severely affected due to both disasters, and return and resettlement formed a major part of the post conflict agenda. A large number of the IDPs were of ethnically Tamil origin, the ethnic group which is claimed to be represented by the LTTE and thereby critically suffered by a war atmosphere. Price (2010) argues that winning their minds and regaining trust were crucial in the post war reconciliation process. However, the wider impact on social cohesion should not be ignored, including those of other minorities such as Muslims.
The end of the war was a great opportunity to distribute the dividends of military cessation throughout the country and achieve sustainable growth (Athukorala and Jayasuriya, 2013). Giving a sustainable solution to those that suffer from physical and emotional displacement is crucial to achieve long-term peace (Price, 2010). Nevertheless, Sri Lanka entered a period of negative peace after war in 2009, a period when physical violence was absent, but there was very little justice (Price, 2010;Zabyelina, 2013). The government which was in power in Sri Lanka, continued to treat IDPs as potential threats and confined them to the camps in highly securitised and poorly facilitated conditions (Price, 2010). Moreover, factors relating to the grievances of the Tamil community and what caused a separatist war, remained unaddressed (Höglund and Orjuela, 2011).
The government that was in power received the majority support of the local population and a high inflow of foreign aid for reconstruction. Yet, it was not able to distribute the peace dividends throughout the country and achieve sustainable development (Athukorala and Jayasuriya, 2013). Reconstruction and economic 573 Tsunami affected communities in Sri Lanka development were viewed as the major tools of peace (Höglund and Orjuela, 2011) while maintaining territorial unity through a centralised approach to power (Price, 2010). Whilst the reconstruction was taking place, the military was expanded in order to stronghold the areas previously controlled by the LTTE (Höglund and Orjuela, 2011). A few major infrastructure projects, including the harbour and airport, were concentrated in the south, the electoral base of the previous President Mahinda Rajapaksa (Athukorala and Jayasuriya, 2013), but also away from the areas of direct engagement between the government and LTTE forces. Nevertheless, reconstruction was undertaken throughout the country, including the north and east, parallel to the resettlement programme and growth incentives in the region (Goodhand, 2012).

Housing reconstruction in post tsunami and post conflict Sri Lanka
Housing reconstruction in post conflict Sri Lanka was challenged mainly by the existing vulnerabilities of poverty, displacement and loss of family members. The context was complex with the high inflow of external assistance and involvement of local and international NGOs, in addition to national and local government stakeholders (Seneviratne et al., 2015). According to Rajasingham-Senanayake (2005) the post tsunami period bought in a heavy inflow of foreign aids as well as heavy interference of international organisations, who had very little understanding of the local culture and customs. Not considering the local capacity and local ownership created new conflicts.
As identified by Brun and Lund (2009), the government typically played a prominent role regarding housing in Sri Lanka, making it a measure of political success and a way of mobilising public support. They concluded that housing was used as a tool of power politics and control over land and that the previous government may have used housing reconstruction and resettlement as a method of controlling the communities, which has resulted in increased tensions. Specifically, Brun and Lund found that people were often resettled in a politically strategic location and people's lives were thus politicised. As the reconstruction activities were trapped with the political intentions, many housing blocks were abandoned. They also suggested that politicisation of reconstruction resulted in unequal resource distribution, which ultimately fuelled further social divisions.
In the aftermath of the tsunami, Brun and Lund (2008) also noted that Sri Lanka did not have a consistent national housing policy. There were professional, ethnic, geographical and cultural biases in housing reconstruction. There was a trend for central government to concentrate on housing development in the south, while leaving the north and east to donors and local governments. According to Nakazato and Murao (2007) there were clearly regional biases in housing reconstruction after the tsunami. Compared to southern areas, reconstruction in north-eastern areas were far below the required level. Seneviratne et al. (2015) also highlighted that Sri Lanka's housing reconstruction lacked a conflict sensitive approach which led to inequalities.
Although there was a large number of studies that examined Sri Lanka's housing reconstruction after the tsunami, there has been a lack of studies on housing in the post conflict context. While some studies have considered housing reconstruction in a conflict environment (Brun and Lund, 2008), they have not considered their impact on conflict prevention after the cessation of armed conflict. Furthermore, these studies were carried out before the end of the war, before formal reconstruction of the north and east took place. The results of reconstruction projects that took place after the end of the conflict are not reflected in these studies. This gap in research on housing reconstruction exists partly because northern and eastern provinces of Sri Lanka became available for reconstruction only after the conflict. Seneviratne et al. (2015) focus on identifying the challenges in addressing housing needs within the context of post conflict housing reconstruction in Sri Lanka. Although this is a recent study, the focus is on challenges of reconstruction, rather than its impact on conflict prevention. This study addresses housing reconstruction that took place as a response to both disasters of tsunami and conflict, particularly in a post conflict environment. It examines the causal links between housing reconstruction and dynamics of conflict prevention, focussing on the housing reconstruction in war affected regions of Sri Lanka.

Results
Four housing projects were selected for study. The first in the outskirts of Batticaloa city and close to the coast. The second, approximately 20 km from Batticaloa, inland. The third project was in Kilinochchi District and the fourth in Jaffna District, both in the North of Sri Lanka. All four studies were undertaken 6-12 months after residents had started to occupy the housing.

Coastal housing project in Batticaloa district
The first project was funded by an overseas government donor agency. Construction work was carried out by contractors from outside the District. The community was a mixed ethnic one of Tamils and Muslims. The housing reconstruction project included 70 houses that were allocated for 20 Tamil families and 50 Muslim families. It also included a number of public facilities, including a mosque, a health centre, a children's park and an internal road system. The location was described by several local residents as the "historical habitat of the Tamil community before the ethnic tensions", and part of the village had a territorial attachment to Tamil community. Politically supported Muslim settlement during war had resulted in an increased Muslim population. Also the allocated land had been "owned by the mosque" prior to the project.
All residents in the settlement were displaced by the war. However, the criteria for selecting Tamil families to occupy the constructed houses was different to that of those for the Muslim families. This decision was made as a result of an agreement between the minister "Hisbulla" and "Pillayan". The two sets of houses, divided across ethnic lines, were separated by a "buffer zone". This layout was agreed by the donor and local planning officer, with no evidence of community consultation. However, residents were not dissatisfied with this separation. Instead dissatisfaction focussed on access to the supporting infrastructure.
The Mosque was relatively larger than the Hindu Kovil, an issue highlighted by many of the respondents. The public "children's park" has been built adjacent to Muslim houses. As a result, Tamil children do not use this park to play.
Respondents from both ethnic groups described increased ethnic tensions, primarily blamed on ownership of the land and the allocation of houses. There remain disputes over land and ownership of houses. People do not have deeds of the houses and part of the land was identified as being owned by a senior political representative.
Many residents complained of excessive political influence during the reconstruction process. The allocation of more housing units to Muslim families together with the pattern of locating houses and other combined facilities, has created ethnic tension 575 Tsunami affected communities in Sri Lanka among two communities with a sense of marginalisation. Common facilities are concentrated in the Muslim housing area, with little attempt to integrate the communities.
Historical tensions among the two ethnic groups have been heightened. The layout of the settlement is not conducive for inter community interaction and the two ethnic groups have little to do with each other.

Inland housing project in Batticaloa District
The second housing project provided houses to 50 families in a village in which all were affected by war and displacement. The project was funded by the Ministry of Resettlement. The criterion for selecting was those families whose houses were "fully damaged". The area has been under the control of LTTE until 2009, and the displacements became high during the last phase of the war. The political influence in taking decisions regarding reconstruction and resettlement had been high throughout the years of the conflict. The houses were provided for a subsidised loan and the ownership is not transferable.
The decisions made with regard to allocation were viewed to be unfair by the beneficiaries. This was widely blamed on poor pre-evaluation of the beneficiaries, which resulted in polarisation of village communethe discrimination is attributed to caste lineage by those who did not receive the benefits. An actual discrimination based on caste is visible in the manner of allocation.
Oral and physical violence has been reported between households. In several instances beneficiaries have already sold the houses for a price higher than the loan to settle the loan, and then moved to other areas. There remains disputes among housing authorities, sureties, new purchasers and initial beneficiaries, due to not settling loan amount

Housing project in Kilinochchi District
This Kilinochchi District project was part of a larger overseas government donor agency housing programme aimed at resettling war affected families in the north. Construction work was carried out by contractors from outside the District. In total, 50 houses had already been completed and some families had moved in. Resettled families were from diverse backgrounds, though all of them were part of the displaced population in the area.
The scheme provided for no common amenities for use by the settlers, except a few wells that supply water for domestic use. There were also no plans to provide such amenities in the future. Many respondents in the community felt there was little opportunity to interact with other residents.
Although the houses were finished and mostly occupied, roads had not yet been paved and electricity supply had not been installed. This had led to widespread dissatisfaction among the residents.

Housing project in Jaffna District
This Jaffna District project was funded by an overseas government donor agency and aimed at resettling war affected families in the north. Construction work was carried out by contractors from outside the District. Approximately 130 families were resettled with a total population of 890 members. The people in these areas are chiefly farm labourers and fishermen. Hinduism is the predominant religion, followed by Christianity which forms the minority group.
The District Secretariat office was responsible for the selection of beneficiaries. Land owners and women headed families were given preference. The number of members in a family was also considered while selecting the beneficiaries. People who are Samurdi beneficiaries were given consideration as these are families whose income is well below the poverty line.
The residents reported no significant tensions between the communities, despite some diversity.

Discussion
Even though the Sri Lankan security forces and the LTTE have been the main parties to the conflict, the tension between community groups at a local level has also been a major feature of the conflict. It is not reasonable to expect such tensions to subside following the end of the war as many of the conditions that led to such inter community tensions are likely to persist even after the war. Conversely, the processes of resettlement and reconstruction has the potential to either minimise or increase tension and conflict among communities, depending on how reconstruction projects are conceived and implemented. The ongoing construction work provides an opportunity to explore how the construction of infrastructure impacts on inter community relations at a regional and community level. Such an exploration will help determine the nature and extent of impact of reconstruction projects on the local population in terms of social inclusion and exclusion, reinforcement of old tensions and the creation of new conflicts, social aspects of infrastructure development, and conflict mitigation.
The focus of data collection was on housing reconstruction projects in the north and east. All these projects were important components of the resettlement and reconstruction process underway in the region, and yet a number of important issues emerged.
There was evidence of dilution from regional development plans to project implementation, thereby undermining reconciliation efforts. In some cases, local political leaders had been instrumental in this process, heightening conflict as a result. Community consultation and engagement was found to be ad hoc, with many beneficiaries not engaged, or neighbouring communities not consulted. Without a space for local involvement in decision making there was found to be a disconnect between the needs of local communities and the post tsunami and PCR projects in place, and little chance to influence processes, policies or strategies.
Three of the projects were identified by respondents as having suffered from both excessive local political interference, and a lack of donor transparency, creating mistrust within and between communities.
The beneficiaries belonged to diverse communities with a post-history of inter community conflict and tension. Conflict was being heighted between ethnic groups, but also within ethnic groups, for example, due to caste discrimination. Many projects have not been planned in such a way as to reduce such conflicts and tensions. Indeed, in some situations, the projects have reinforced, rather than reduced them. Some schemes had no significant provision for shared infrastructure, while others had infrastructure that was located or designed in such a way that it discouraged use by some groups within the community. Such planning decisions were not always unwelcome by the residents, but donors and local officials also appeared to abdicate their wider responsibilities for promoting conflict sensitivity and peace building in the planning of housing schemes and promotion of shared infrastructure.

577
Tsunami affected communities in Sri Lanka In most cases, contractors came from outside the communities where the projects are located. They usually had their own supply chains and workers brought from outside the region. Many local people felt they did not have opportunities to engage in construction work and gain experience and economic benefits. Three of the projects had been undertaken by inexperienced project teams without appropriate skills to tackle sensitive issues and disputes.

Conclusions and recommendations
Using four housing reconstruction projects in Batticaloa, Kilinochchi and Jaffna Districts, Sri Lanka, as case studies, and a novel methodological framework, this study has explored the causal relations among the independent variables associated with housing reconstruction and dependent variables related to conflict prevention. The data, gathered from interviews and project reports, were analysed using propositions from a literature review, adopting a thematic analytical approach. This study examined post disaster housing reconstruction in Sri Lanka in a post tsunami and post conflict environment, a subject not adequately addressed by previous studies, which have tended to focus on the reconstruction in the immediate aftermath of the tsunami.
The results reveal concerns in infrastructure development surrounding marginalisation of beneficiaries, segregated infrastructure that reinforces divisions, inadequate consultation with target population, and a lack of economic opportunities for local people despite large scale construction activity. The resultant housing and associated infrastructure is often not meeting the needs of marginalised groups, and can create or heighten tensions within and/or among ethnic or religious groups.
While it is apparent that in the short term, provision of basic housing and infrastructure for social protection is important, in the medium to long term a more socially inclusive and equitable development process will be vital to promote reconciliation and co-existence. There remains little understanding or monitoring of the socio-economic impact of infrastructure projects. Many stakeholders recognise the importance of infrastructure in sustaining long-term peace, and the politically sensitive nature of any attempts to highlight such issues. There is a need to further sensitise key stakeholders regarding the principles of socially inclusive and equitable infrastructure development, including donors, national and local authorities, and contractors. There is also a need to develop and institutionalise grievance redress mechanisms for marginalised and vulnerable groups such as ex-combatants, youth, women and disabled, and to monitor and evaluate the social and economic impact of infrastructure development. There is also a clear need to increase market access for the north and east construction industry, including local entrepreneurs and labour.