Participation in Humanitarian Aid: Sri Lanka report August 2002

Transcription

1 Consultation with and Participation by Beneficiary and Affected Populations in Planning, Managing, Monitoring and Evaluating Humanitarian Aid: the case of Sri Lanka Prepared for INTRAC by Jo Boyden with Tania Kaiser and Simon Springett and commissioned by ALNAP August

2 Acknowledgements This report is based on nine weeks fieldwork in Sri Lanka between February and April During this period the authors benefited immeasurably from the skilled translation, assistance, guidance, enthusiasm and friendship of the research team. We are deeply grateful to F. Felician, Mark Paterson, Umesh, Tharmila, Jude Simion, Anberiya Hanifa, Dr. Perinpanathan, S. Inparuban, K. Mahendren, U. Sangaralingam, M. Thushyanthan, and T.T. Mayuran. Although the authors must take full responsibility for the contents of the report and any possible errors therein, many people kindly read and commented on the various drafts. We would like to thank Alastair Kirk, Jeevan Thiagarajah, Patrick Vandenbruaene, Jason Hart, Brian Pratt, David Marsden, Phil Esmonde, Nicola Mushet, Raga Alphonsis, Jonathan Goodhand, Shah Liton, Ravi Kumar, André Griekspoor and the ALNAP Steering Group and Secretariat for their very detailed and constructive feedback. In Sri Lanka we were hosted in Jaffna by the Save the Children (UK) and in Vavuniya by Oxfam-GB. Oxfam-GB also arranged our security clearance and visas. We owe a debt of gratitude to these agencies and to their staff in both Colombo and the research sites for their hospitality, logistical support and general forbearance over the many demands we made on them. The Consortium of Humanitarian Agencies in Sri Lanka generously hosted a dissemination workshop in Colombo at which the preliminary findings from the research were shared. This event was extremely helpful to us in providing a general framework for the subsequent report. The case study was set up and organised by INTRAC and the agency provided consistent intellectual and administrative support throughout. In this regard, special thanks are due to staff members Michael Davis, Jerry Adams, Brian Pratt and David Marsden and to INTRAC consultants Tony Vaux, Peter Loizos and Jonathan Goodhand. The Global Study, of which this report is but one component, was managed initially by Peter Oakley, then research director at INTRAC. Due to his tragic and untimely death, Peter ran the project for only a short while during its initial stages. However, his leadership, vision and early insights remained an inspiration to us all throughout. 2

3 List of Acronyms Executive Summary Table of Contents Part I. Background 1. Introduction 1.1. Participation in humanitarian action 1.2. The Global Study 2. The Sri Lankan Case Study 2.1. Conceptual issues 2.2. Methodological issues 2.3. Ethical concerns 2.4. The choice of field study sites 3. The structure of this report Part II. The Operational Context 1. Country Profile 1.1. History of the conflict 1.2. Civilian experiences of conflict 1.3. Humanitarian responses 2. Operational constraints 2.1 Political intervention in humanitarian action 2.2. Psychological and emotional obstacles 2.3. Social and cultural obstacles 2.4. Negative perceptions of participation & humanitarian aid 2.5. Indifference of camp populations 3. Summary and observations Part III. Participation in Practice 1. Participation in the project cycle 1.1. Planning 1.2. Baseline assessment 3

4 1.3. Project Implementation 1.4. Monitoring and evaluation 2. The benefits of participation 2.1. Self-respect and self-efficacy 2.2. Social skills and decision-making 2.3. Self-representation, self-protection and leadership 2.4. Solidarity and empowerment of vulnerable groups: women 2.5. Solidarity and empowerment of vulnerable groups: children 2.6. A sense of ownership and the material rewards 3. Summary and observations Part IV. Mechanisms for Promoting Participation 1. Participation strategies 1.2. Deciding on the operational model 1.2. Breaking down the barriers 1.3. Social mobilisation 1.4. Change agents 1.5. Use of culturally approved idioms 1.6. Social targeting 2. Promoting participation through institutional means 2.1. Informal interaction and exchange 2.2. Formal organisations 2.3. Local and national partner organisations 3. Funding and participation 3.1. The potential of donors to promote participation 3.2. Detrimental aspects of funding 4. Summary and observations Part V. Recommendations Notes 4

5 5

6 List of Acronyms Used ALNAP ACF CBO CHA CIDA DfID DRC EHED EPDP EPRLF ESCO GDP GS GTZ ICRC INTRAC IPKF LTTE MoU MSF NHDA NFRI PRA QIP RDF SEDEC SLA TRO TELO TNA TRRO YMCA Active Learning Network for Accountability and Performance (in Humanitarian Assistance) Action Contre la Faim Community based Organisation Consortium of Humanitarian Agencies Canadian International Development Agency Department for International Development Danish Refugee Council Eastern Human Economic Development (Local branch of CARITAS) Elam People s Democratic Party Elam s People s Revolutionary Liberation Front Eastern Self-Reliant Community Awakening Organisation Gross Domestic Product Grama Sevaka German Technical Co-operation International Committee for the Red Cross International NGO Training & Research Centre Indian Peace Keeping Forces Liberation Tigers of Tamil Eelam Memorandum of Understanding Médecins Sans Frontières National Housing Development Authority Non-food relief items Participatory Rural Appraisal Quick Impact Projects Rural Development Foundation Social and Economic Development Centre (Caritas Sri Lanka) Sri Lankan Army Tamil Rehabilitation Organisation Tamil Elam Liberation Organisation Tamil National Army Tamil Relief and Rehabilitation Organisation Young Men s Christian Association 6

7 Executive Summary Part I The participation of populations assisted by international humanitarian action in measures supporting them is now widely accepted as crucial to effective social targeting, resource utilisation, accountability, sustainability and impact. For some, participation is also a fundamental right of citizenship, essential in the context of humanitarian emergencies for survival, self-protection and self-actualisation. Despite clear institutional commitments to consultation and participation at the policy level, however, there remains wide variation in practice. Indeed, there is increasing concern over the lack of consultation with and participation by disaster-affected populations in the design, management, implementation and monitoring and evaluation of relief programmes. It is against this background that ALNAP commissioned a Global Study on the subject. The intention is to reveal the mechanisms by which the voice of affected populations can be enhanced within the humanitarian system, while remaining alert to the potential difficulties in engaging in such a process in emergency contexts. The present report, which draws on primary (aid recipient) and secondary (agency) stakeholder perceptions in Sri Lanka, is the pilot in a series of five country case studies that provide the empirical evidence for the Global Study. Testing the hypothesis that the active consultation and participation of crisis-affected populations in measures to assist them is (according to aid recipients and other key stakeholders) both feasible and beneficial in terms of project outcome, it investigates current policy and field practice in three locations affected by conflict in the north and east of the island. Part II The ethnic conflict in Sri Lanka between the Tamils and Sinhalese has varied in intensity and location and has been marked by a number of major episodes of severe violence. It has been experienced very differently by different sections of the population in the north and east, although displacement is a major feature of civilian life in all three of the areas where research was carried out. Humanitarian actors, local, national and international, have provided relief to camp dwellers for many years, but a growing number are now focusing also on rehabilitation and development activities in resettlement and relocation villages. Most humanitarian actors have a theory about why the participation of aid recipients in measures supporting them is beneficial and valid. Such theoretical perspectives determine the objectives, strategies and outcomes of humanitarian programmes and projects. In Sri Lanka two theoretical models predominate. The majority of agencies follow what here is termed an instrumental approach, in which the prime objective of involving aid recipients in the various activities of a project is to improve overall project performance. In other words, participation is viewed as a means towards a specific end, which is the delivery of effective projects. A far smaller number of 7

8 agencies adhere to more transformative models, in which aid as regarded as contributing to fundamental societal change. Such approaches seek to empower aid recipients to assume greater control over their lives, undertake action to improve their condition and circumstances, play an active role in political and civic processes and forge relations with others that contribute to peace and to the development and enhancement of the wider community. Many agencies are committed to beneficiary participation at the policy level but find that their efforts are thwarted in practice by a variety of constraints. Some of these constraints have to do with the security and political pressures associated with the conflict, some arise out of contrasting social and cultural values, and some derive from psychological, emotional and economic difficulties confronted by civilians or from their negative perceptions of humanitarian aid. To emphasise the many constraints to participation that predominate in war-affected areas and to highlight the differences in social and cultural ideas on the subject is not to try and dissuade humanitarian agencies from consulting and developing participatory programmes with their clients. It is intended more to alert them to the need to reflect on the potential and actual limitations and risks associated with participatory approaches. It is also intended to stress the importance of canvassing both primary and secondary stakeholder perceptions on the value added and possible dangers of participation and of engaging in regular and continuous risk assessment and project monitoring and appraisal. The factors mitigating against the more empowering and transformative models of participation in particular may be so great that it may be unrealistic for agencies to aspire to such radical goals. Humanitarian measures may need to observe more modest objectives, whilst always taking every opportunity possible to consult and inform their clients and other members of the affected population. Part III The degrees and forms of beneficiary participation vary considerably according not just to environmental opportunities and constraints but also to the objectives and underlying theoretical perspectives of humanitarian agencies. Such variations can be observed in the different stages and activities of the project cycle. Most of the agencies in Sri Lanka have adopted an instrumental view of participation that favours consulting beneficiaries (especially during baseline assessment) about their views, problems and needs, informing them and building in them the commitment and competencies necessary for active engagement in project implementation. Because project design and approval does not involve beneficiaries directly, however, most find it difficult to offer aid recipients a meaningful role in planning. This sets a precedent for engagement with beneficiaries which is not very conducive to sustaining high levels of participation throughout later stages of the project cycle. While beneficiary participation in planning is low, greater effort is made to foster participation in implementation. This commonly entails the contribution of beneficiary ideas, knowledge, labour and other skills to construction, maintenance and administration of project resources. Project monitoring during implementation is fairly informal, although in many cases it is quite regular and in some cases quite 8

9 participatory. In general, beneficiary enthusiasm for projects tends to diminish towards the end of the cycle and few agencies engage in participatory impact evaluation, even while most recognise its importance. Examples of more transformative approaches in which clients are empowered to represent themselves before the authorities, engage in autonomous collective action on community and other issues and assume control of project resources are comparatively rare. However, some projects embody certain elements of this kind of approach and some organisations are moving gradually in this direction as confidence in the cease-fire grows, more developmental approaches are being employed and civilians begin to return to their communities. It is evident that blueprint projects are not very effective in terms of fostering high levels of participation throughout the project cycle. It is also apparent that greater flexibility and openness to beneficiary involvement in the earlier stage of the cycle and in project appraisal would make a difference in this regard. But, given the political and other constraints, it is not clear that this kind of innovation is possible in many cases in the Sri Lankan context. Also, based on the reflections of a significant proportion of respondents, such innovation is not always sought by aid recipients. Despite the many limitations, there is evidence that aid-recipients can and do benefit by being better informed and consulted and by assuming meaningful roles in project implementation. These benefits include the development of important personal and social skills and a greater sense of ownership and pride in project outcomes. Overall, targeted projects with women on the one hand and with children and youth on the other have been more successful in terms of fostering beneficiary participation than more generic projects affecting whole population groups or projects with men. They have, according to the reports of aid recipients, also had more beneficial impacts. Of all groups, children and young people who have been involved in psychosocial programmes report the most radical effects. This may be because humanitarian measures normally take place in communities whilst men are at work elsewhere, or it could be due to inordinately high levels of alcohol use among men in the north and east. Part IV Levels of beneficiary consultation and participation in humanitarian aid have in the past often been very low or non-existent and such ideas and practices are not very developed in many of the more hierarchical and prescriptive cultures of the world. This implies that humanitarian actors that are committed to the approach must develop explicit means for promoting participation. These means are used in varying ways, degrees and forms depending on the model of participation adopted. In Sri Lanka humanitarian agencies are heavily reliant on locally recruited change agents to mobilise village and camp inhabitants to take an active part in project implementation. The orientation and training of these animators is of critical importance to project outcome, as is their support. Many agencies regard the recruitment of animators and mobilisation of beneficiaries as a first step in creating a more formal institutional structure (commonly a CBO) within which collective social action can be promoted and directed. These structures vary in the extent to which they become truly self-governing, autonomous entities and often in practice they remain 9

10 very dependent on the implementing agency. Management of CBO funds by implementing agencies places very real limits on beneficiary empowerment and selfefficacy. With the increased reliance on partnerships with local and national organisations, the lengthening of funding cycles and development of more open and flexible relations between donors and implementing agencies there have been some very promising trends in institutional relations in recent years. To this we should add efforts in some quarters to create less authoritarian management structures and styles and improve aid co-ordination. Certain donors have played an important role in this respect, encouraging implementing partners to avoid duplication and collaborate with each other whenever possible. These developments have considerable potential to further promote beneficiary participation. Nevertheless, there remain serious problems in some sections of the aid community in terms of political intervention in humanitarian action, high turnover in staff, inflexible and short funding cycles, competition over beneficiary populations and conflicting aims and strategies. In many cases, humanitarian actors are party to major decisions and policies that have a crucial impact on the lives of beneficiaries and yet make no effort to consult or keep affected populations informed. Such trends have the capacity to undermine agency: beneficiary relations. Part V On the basis of our findings we would suggest that consultation and the timely provision of accurate information on critical issues is greatly valued by beneficiaries and is a must in all circumstances. From the point of view of aid recipients, consultation and information provision are important instruments through which agencies demonstrate respect for their clients. They are also the means by which beneficiaries can gain a greater sense of control over their lives. Assuming that the environmental conditions are sufficiently conducive, aid implementers have the appropriate skills and capacity, donors are supportive and aid recipients view the idea favourably, more active and meaningful participation is also feasible within the field of humanitarian action. However, it has to be recognised that the employment of more participatory approaches presents many challenges. For a list of specific recommendations on how to increase beneficiary consultation and participation please refer to Part V. 10

11 I. Background 1. Introduction 1.1. Participation in humanitarian action The participation of populations assisted by international humanitarian action i in measures supporting them is now widely accepted as crucial to effective social targeting, resource utilisation, accountability, sustainability and impact. For some, participation is also a fundamental right of citizenship, essential in the context of humanitarian emergencies for survival, self-protection and self-actualisation. As such, client participation has become a central tenet of policy for a number of humanitarian agencies globally, is incorporated into many mission statements and in some cases is constitutionally enshrined. ii Thus, the introduction to the key 1994 Red Cross/NGO Code of Conduct commits signatory agencies to involve programme beneficiaries in the management of relief aid. The World Food Programme provides that it will ensure that its assistance programmes are designed and implemented on the basis of broad-based participation. iii Likewise, in a section of UNHCR s Handbook for Emergencies devoted to community involvement, it is stated that: refugees must be involved in planning measures to meet their needs and in implementing those measures. The way the community is organised can help ensure that the refugees skills are made use of and that the personnel for services at the site will come from the refugees. iv The trend toward increased participation in humanitarian interventions is underpinned by a growing recognition that beneficiaries of aid are not merely the passive recipients of assistance but social actors with insights into their situation and competencies, energy and ideas that can be harnessed to improve their circumstances. More than this, they are primary stakeholders in humanitarian action with significant vested interests in humanitarian policies and interventions. They have views about possible solutions to their problems and understandings of how humanitarian measures can support broader societal objectives of peace, reconstruction and development The Global Study Despite clear institutional commitments to participation at the policy level, there remains wide variation in practice. Indeed, there is increasing concern over the lack of consultation and participation by disaster-affected populations in relief projects and programmes. It is against this background that ALNAP, an international network of humanitarian agencies, commissioned a Global Study entitled Consultation with and Participation by Beneficiary and Affected Populations in Planning, Managing, Monitoring and Evaluating Humanitarian Aid. The present report on Sri Lanka is the pilot in a series of five country case studies that provide the empirical field data for this project. The core objectives of the Global Study are: 11

12 To assess current consultation & participation practice in a range of emergency contexts; To identify examples of good practice; To identify gaps or inadequacies in current practice and contributing factors; To improve understanding of participation and consultation practice. The intention is to reveal the mechanisms by which the voice of affected populations can be enhanced within the humanitarian system, while remaining alert to the potential difficulties in engaging in such a process in conflict contexts. Thus, affected populations are situated at the centre of the study, which seeks to understand how they perceive and interact with the myriad of governmental, international, national, local and other institutions that manage, regulate, control and influence the delivery of humanitarian assistance and protection. v Wherever possible, successful consultative and participatory mechanisms and initiatives whose use could be more widely promoted in the interests of greater programme effectiveness and downward accountability are identified. The five country case studies will cover a broad spectrum of humanitarian responses, including those established in the context of long-term complex political emergencies and those following rapid-onset environmental disasters. The project will yield several outputs: in addition to five country monographs, an overview volume will synthesise the key learning points from these studies and a guide on client participation in humanitarian action will be produced for practitioners. 2. The Sri Lankan Case Study Sri Lanka was chosen as a case study country by ALNAP because some relevant studies had already carried out there and these could be used for background and contextual information. It was selected as the pilot for the Global Study on the grounds that a wide diversity of humanitarian agencies have been present in the country for many years. These agencies have mounted a range of relief, rehabilitation and development interventions with war-affected populations in camps and resettlement and relocation villages, providing considerable scope for investigation of the topic of participation and development of the research methodology and methods Conceptual issues There is a major literature on client participation in aid interventions invoking a wide variety of interpretations and definitions. In this study participation is understood in its simplest sense to imply the active engagement of primary stakeholders in the planning, management, implementation and assessment of humanitarian measures affecting them. The various facets of participation include information sharing, consultation, the contribution of manual labour and other skills, involvement in decision-making and/or resource control. These facets of participation are often taken to represent increasing gradations of engagement in humanitarian measures, as follows: Information sharing; minimally informing affected populations about measures and decisions affecting them; Consultation; some level of consultation with stakeholders within programme 12

13 guidelines; Contribution of labour and other skills; input of stakeholders into construction and other forms of project implementation; Decision-making; direct involvement of affected populations in decisions made during the project cycle; Resource control; control over project resources assumed by the client population and all the major decisions over these resources and over any new initiatives are made by clients. vi We note that the study treats consultation and participation as independent but interrelated dimensions of engagement in humanitarian measures. While consultation with beneficiaries is an important aspect of participation, consultation can also occur independently in projects that do not have participatory goals and objectives. vii Testing the hypothesis that the active consultation and participation of crisis-affected populations in measures to assist them is (according to aid recipients and other key stakeholders) both feasible and beneficial in terms of project outcome, the Sri Lanka study investigates current policy and field practice in different field settings. Given the divergence in agency perspectives on participation in Sri Lanka (see Part II, section 1.2.), one of the key challenges of the field research was to compare and contrast the prevailing theories and models to establish both their suitability and impacts in the context of a long-term conflict Methodological issues This report draws primarily on qualitative data derived from primary stakeholder (beneficiary) and secondary stakeholder (agency) perceptions and to a far lesser extent from secondary sources. The literature on Sri Lanka, the conflict and responses to it, was consulted and a range of project documents reviewed. Interviews were conducted in Colombo with representatives of humanitarian agencies and with key academics and consultant practitioners in humanitarian action. The field research was carried out in three locations. General meetings with agency personnel were held in two of these locations and individual interviews were conducted with agency staff in all three. Some or all of the team met and held interviews and focus group discussions with representatives from a broad range of villages and camps in war-affected areas. Care was taken to ensure that men, women, children and members of different social and economic groups were included in the research. Where views differed, this is indicated in the body of the report. Drawing mostly on accepted Participatory Rural Appraisal (PRA) tools, several group-based research methods were employed during the research with crisis-affected populations. These were complemented by methods, mainly unstructured and semistructured interviews, which were employed with individual respondents and by group discussions and observation. Focus group discussions emerged as the most productive way to canvass views in a public context. One-to-one interviews provided an opportunity for triangulation and for discussion of issues that were too sensitive or complex to be raised in a public forum. 13

14 2.3. Ethical concerns Participation in Humanitarian Aid: Sri Lanka report August 2002 Many of the civilians we spoke to in the north and east of Sri Lanka have experienced multiple violations and losses. There can be considerable therapeutic value for survivors of violence and armed conflict in giving testimony of their experiences, especially if they do so in a safe environment. Respondents were at times extremely frank with us. However, there were many subjects that we felt unable to explore for fear of jeopardising people s safety or causing further distress. For the same reason, most of the information on specific incidents and violations has been excluded from the report. We were also very concerned that many people seemed to assume that we were human rights monitors with the ability to bring perpetrators of violations to justice. We repeatedly stated that this was not our function, but sometimes felt that our actual role had not been well understood. The presence of foreigners in the research team led some respondents to believe that we were agency representatives undertaking an assessment prior to delivering a good or service. This was exacerbated by the fact that many of the field visits in villages and camps were attended by a representative of an assistance-providing organisation, who sometimes also introduced the research team to the respondent group or played a facilitation or translation role. All possible efforts were made to avoid raising expectations, but the team was worried about the extent to which this remained an issue in some of the communities visited. viii Further, while every effort was made by all concerned to maintain as much neutrality as possible, the presence of agency personnel inevitably had an impact on the data gathered The choice of field study sites This case study focuses on humanitarian action in the context of a long-term, intrastate conflict. Three sites were selected for inclusion in the study: Batticaloa in the east of Sri Lanka, and Jaffna and Vavuniya/Mannar in the north. They were chosen because they have been exposed to major outbreaks of violence, represent a broad range of military, political and socio-economic contexts and have long been centres of humanitarian action. Choice of these sites enabled inclusion in the research of areas that are controlled by the government (termed by many cleared areas ), areas controlled by the LTTE ( uncleared areas) and grey areas that are controlled by government by day and the LTTE at night. They also encompass border areas, which are on the front line or on the line dividing different ethnic and/or religious groups. The majority of the population in these regions is Tamil, although there are significant Muslim populations in some areas of Batticaloa and a small number of Sinhalese in Vavuniya and Mannar in particular. The one major war-affected region that was excluded from the study is the LTTEcontrolled Vanni. It was decided that this would have proved very difficult as a field site because access and mobility are highly restricted and the possibility of doing participatory research there is very limited. The consequences of not working also in areas such as Trincomalee and Anuradhapura are that the study focuses principally on Tamil, and to a lesser extent, Muslim populations and does not include those Sinhalese people who have also been displaced or otherwise affected by conflict. There has, of course, been arguably related conflict in Sinhalese areas of the south of the country. These areas were not included in the study because there is less 14

15 humanitarian activity in the south at present than in the north, and such activity as there is tends to be focused more on the environmental as opposed to political emergency. The Batticaloa Field Study focuses on the work of Eastern Human Economic Development (EHED ix ), the YMCA and a range of local and national NGOs, including Sarvodaya, Suriya and the Women s Development Forum. These agencies work mainly in resettlement and relocation villages and engage mostly in rehabilitation and development programmes, the scope of which is extremely broad and covers micro finance, pre-school education, water and sanitation and many other measures besides. The Vavuniya/Mannar field study concentrates largely on the work of Oxfam-GB and some of its partner organisations (especially Sewa Lanka, RDF and SEED) and to a lesser extent on FORUT, UNHCR, the Danish Refugee Council and ZOA Refugee Care, Netherlands. The projects reviewed include relief initiatives (mainly services and non-food relief items) in both IDP/returnee-refugee camps and a wide variety of rehabilitation and development measures in resettlement and relocation villages. The Jaffna field study focuses principally on the work of Save the Children (UK), and to an extent on the programmes of CARE International and UNHCR. A broad mix of relief, rehabilitation and development initiatives in both IDP camps and war-affected villages were reviewed in this area. 3. The structure of this report The findings from the field research and their analysis are provided in Parts II-IV and the recommendations are given in Part V. Parts II-IV all finish with a brief summary and observations. These outline the main issues raised in the text. Given the importance of understanding how the history of an emergency and the local context more generally influence participatory efforts, Part II offers an overview of the operational environment within which humanitarian agencies operate in Sri Lanka, highlighting in particular some of the obstacles encountered. Part III focuses more on implementation and begins by exploring the various forms and levels of participation that commonly prevail during the various stages of the project cycle. Following this is a brief discussion of some of the most significant benefits associated with participation, many of which were identified by aid recipients themselves during interviews and focus groups. Part IV details the main mechanisms and strategies, operational, institutional and financial, that have been developed by agencies in Sri Lanka as a means of promoting participation and Part V provides recommendations for both implementing and donor agencies. 15

16 II. The Operational Context Part II begins with a brief history of the conflict in Sri Lanka and of the humanitarian response. This is followed by a more detailed account of the key constraints to the implementation of participatory approaches. 1. Country Profile 1.1. History of the conflict Numerous reports on the Sri Lankan ethnic conflict exist in both published and grey literatures and hence an exhaustive account is unnecessary here. x The conflict has its origins in the 1930s when certain sections of the Tamil speaking community (around 19% of the total population according to the census of 1981) were over represented in the colonial administration. xi Following Independence in 1948, the state has been dominated by the majority Sinhalese population. Over the years the government has introduced a number of discriminatory measures limiting Tamil political representation, curbing access to education and land and inhibiting the use of the Tamil language. For their part, the Tamils have consistently sought a share in power within the legislature, an aim that is refuted by the Sinhalese. The failure of successive Sinhalese-dominated governments to resolve their political differences with the Tamil leaders has led to a marked deterioration in relations between the two populations. With time, the demands of the Tamil community have become more fundamental and the means adopted by militant groups - of achieving these demands, more violent. The present claim is for a separate state, Tamil Elam, in the north and east of the island. The ethnic violence has been at its most extreme in the last two decades. The 1983 communal riots against the Tamils, which many see as the critical turning point in the conflict, led to full insurgent warfare. Several militant Tamil groups emerged in the north and east in opposition to the Sri Lankan Army (SLA) and there followed a process of continuous repression and unrest with three major periods of mass violence, known as the Elam Wars. India was heavily involved in brokering a settlement to the first of these conflicts, instigating the 1987 Indo-Sri Lanka Agreement, and the Indian Peace Keeping Forces (IPKF) acted as mediators, arriving in the country the first time in The involvement and conduct of the peacekeepers has been interpreted variously, but there is evidence that they were far from neutral politically and perpetrated many serious violations against Tamil civilians. The LTTE, which now claims to be the sole representative of the Tamil population, engaged the IPKF in a 2-year war of attrition. The India-Lanka Accord eventually collapsed and the IPKF withdrew from the country in During the brief peace that followed the government and the LTTE held peace talks. However, these collapsed and the second Elam war ( ) broke out, with the Sri Lankan army and Tamil militant groups (which in their opposition to the LTTE, had switched sides) 16

17 fighting on one side and the LTTE on the other. Another short period of peace followed with the election of a new government that sought resolution of the conflict. But negotiations between the warring parties collapsed in April 1995 and Elam War 3 broke out. Apparently committed to bringing an end to the conflict, the UNP gained a victory in the 2001 elections and has subsequently received further electoral endorsement for its attempts to find a political settlement to the conflict from the Sri Lankan people. The armed confrontation between the LTTE and the government of Sri Lanka may have been intermittent, but it has lasted for almost twenty years, cost an estimated ,000 lives xii, and caused the multiple displacement (prior to the cease-fire) of around 800,000 people. xiii The economic costs of maintaining the armed forces as well as bolstering systems of public security has been extremely high, the percentage of GDP allocated to defence having grown from 0.5% of GDP in 1982 to 3.11% in 1992 and 5.3% in xiv This, in turn, has caused a further drain upon resources needed for the provision of basic government services and for the development of the economy and the country s infrastructure. The disparity between defence and social spending by the Sri Lankan government is marked. Throughout the conflict, Tamil, Sinhalese and Muslim ethnic and religious identities have been manipulated by various leaders for political ends. xv Yet, contrary to popular stereotype, the conflict has not produced a clear and straightforward division between groups. For one thing, the Muslim community has also been profoundly affected, although normally this is little acknowledged. For example, an estimated 100,000 Muslims were displaced from the north following an LTTE ultimatum in October 1990 ordering them to leave the region and many still remain in camps in Puttalam and Anuradhapura districts. xvi Second, there is a sizeable Christian minority in Sri Lanka that is composed of both Sinhalese and Tamils, blurring the division between the two. Third, while the LTTE has gradually emerged as the major political force within the Tamil community, there remain significant political distinctions and disputes within this population. And the conflict in the South showed that the Sinhalese are also seriously divided. Thus, stark distinctions based on ethnicity often mask internal political divisions and caste/class differences that are extremely powerful and have the potential to seriously undermine the participatory efforts of the humanitarian community. The present cease-fire (which came into force in December 2001), the signing of the cease-fire agreement on the 23rd February 2002 and pending talks between the government and the opposing forces, the Liberation Tigers of Tamil Elam (LTTE) have brought about major changes in the lives of civilians and in the operational environment for humanitarian actors. xvii Nevertheless, at the time of field research Sri Lanka could by no means be described as post-conflict. xviii 1.2. Civilian experiences of conflict in the research sites Even though civilians throughout the island have all been affected by conflict in one form or another, civilians in the north and east of Sri Lanka have been more directly affected than others by the ethnic violence between Sinhalese and Tamils. Large numbers in these areas have lost their lives or disappeared. Large numbers have experienced multiple displacements over a long period of time and in many cases 17

18 communities and families have become separated and widely dispersed both within Sri Lanka and overseas. For those settlements subjected to frequent military incursions and shelling, displacement may occur regularly on a weekly or even daily basis. Many families have become accustomed to living in their villages by day and retreating to the forests at night. Quite apart from displacement, many civilians have been exposed to horrific violations against the person, ranging from sexual and gender violence, abuse and exploitation, to disappearances, torture and extra-judicial killings. Many have become impoverished due to mobility restrictions, prohibitions on access to cultivable lands and fishing grounds and to overall economic collapse. For the purpose of understanding in broad terms the circumstances of civilians living in the north and east, the population of these areas can be said to fall into five broad categories: 1) those in their original homes and communities who were never displaced; 2) those who are displaced and living in camps (including internally displaced and those refugees who have returned from camps in India); 3) those who are displaced and self-settled with friends and/or relatives; 4) those who have returned to their original communities (termed resettlement villages) following displacement; 5) those in relocation villages, who have left camps and have been relocated to new settlements, primarily within the district they originally came from. Civilians in all five categories were included as respondents in the study. Of the three study sites, Batticaloa District is the poorest and contains both Tamil and Muslim populations. Up to 80% of the people live in the government-held area along the coast, while only 20% (all Tamils) live in the interior in LTTE territory. Although civilians have experienced periods of extreme violence in the past and continue to be subject to harassment, kidnapping and forced recruitment by political and security bodies, the district is of less strategic interest than the other two areas and hence is more settled in many ways. In Batticaloa most families are in their original communities or have now returned to their homes following displacement, although some remain displaced and are self-settled locally. Quite a few have migrated overseas and a very small number remain in camps. As an area of critical strategic importance, the Jaffna peninsula has been subjected to major military incursions and waves of large-scale displacement. The present population of around 500,000 is thought to be about half of the original number of inhabitants and many people have fled elsewhere on the island or overseas. The peninsula is now under government control and certain high security areas remain out of bounds to civilians. Overall, the region is heavily contaminated with landmines and unexploded ordnance. A significant number of people in Jaffna have now returned to their original homes and agricultural production and farming is gradually recovering in some areas. People who remain in camps in Jaffna tend to be those originating from high security zones that are presently controlled by the SLA. The possibility that they will return to their villages within the foreseeable future is very slim. Situated next to the LTTE heartland of the Vanni and on the route to Jaffna, the original home of Tamil resistance, control of Vavuniya has now been largely wrested 18

19 from the LTTE by the Sri Lankan armed forces who maintain a heavy presence in the area. The population is more ethnically diverse than the other two regions and includes Sinhalese, Tamil and Muslim communities. The first camps in Vavuniya were created in 1992 to accommodate refugee returnees from India. Among the displaced population of the region many are up country Tamils who originally came from the tea estates in the central highlands of Sri Lanka. In 1994 the camps received a new influx of displaced people from Jaffna and the Vanni. There are now 17 camps in the area that in March 2002 housed a reduced population of approximately 15,000 individuals xix Conditions in the camps, now controlled by the government, are very poor. In many cases they have deteriorated since the cease-fire because the government endorses resettlement and relocation and opposes further investment in camps. A significant number of displaced people in the area have been settled by the government in grey and border areas within the district where they are vulnerable to the fighting Humanitarian responses to the conflict The critical trends in humanitarian aid A number of international humanitarian agencies have been present in Sri Lanka since the onset of fighting and many of the local and national organisations that were engaged in development provision prior to the conflict subsequently became involved in humanitarian aid. The diversity of humanitarian programme types and approaches is great and this was reflected in the interventions reviewed by the research team. While some of the projects investigated can be clearly defined as emergency relief, others focused more on reconstruction, rehabilitation, or development in support of resettled populations and others exposed to protracted political conflict. The operational context for humanitarian measures in Sri Lanka has been highly charged. Humanitarian actors of all kinds have pointed to the delicacy of a situation in which they have been perceived by some as responding to humanitarian needs caused deliberately by warring parties as part of their politico-military strategies. A commitment to participation by recipients of aid is apparent in many quarters, although, as we argue in the report, in this kind of setting many agencies have found it quite difficult to realise such a commitment in practice. Since the outset of the gravest fighting in the mid-1980s, the humanitarian community has tried to implement its programmes in line with the principle of neutrality. This has meant supporting populations in government-controlled areas and bringing aid items in via these areas (transferring them at the Forward Defence Line) to LTTE-controlled areas where they are distributed to the local administration, international and local agencies. xx In the early 1990s the LTTE began to increase its relief capacity through the work of the Tamil Rehabilitation Organisation (TRO), which is administered by its political wing. With time, the TRO has exerted increasing pressure over and constraints on international and local relief organisations, to the point that prior to the cease-fire, many had effectively retreated to government-controlled and grey areas. Although most agencies acknowledge that the situation of populations in LTTE territory is in many ways worse than that of people living in government areas, they argue that low levels of support to the former are attributable to unreasonable restrictions imposed by the LTTE. 19

20 Institutional responses to humanitarian need have waxed and waned as the intensity and location of violence has shifted and the overall approach has altered markedly over the prolonged period of conflict. The following broad trends can be identified: A clear historical division between development aid in the south and humanitarian aid in the north and east of the island has curtailed beneficiary consultation and participation in the latter on the grounds that this part of the country is subject to war. However, this outlook is gradually changing; A movement away from relief interventions towards rehabilitation and development (termed development-relief by some) with war-affected populations; A transition from direct implementation by international organisations to working through local and national partner organisations; A reduction in the deployment of expatriate staff to field offices; A move by some agencies from short-term to longer-term funding cycles; Increased efforts to co-ordinate humanitarian programmes; A growing recognition (by some agencies only) of the importance of client participation and international humanitarian standards; Increased utilisation of a human rights framework and support for advocacy in addition, or as opposed, to direct service provision; A shift in attention away from IDP/returnee camps and towards support for resettlement and relocation villages; An increased focus on livelihood security (especially micro-finance) and physical infrastructure in resettlement and relocation villages in particular. Agency commitments to beneficiary participation Among humanitarian agencies in Sri Lanka there is a fairly high level of acceptance of the concept of participation. Many of the agencies in the country that are involved in humanitarian action have in varying forms and degrees incorporated beneficiary participation as a major facet of their work. xxi One respondent xxii argued that the humanitarian mandate, which he regards as being in essence quite restrictive, compels agencies to introduce participatory approaches. Humanitarian aid in conflict and immediate post conflict settings tends to entail short-term projects that focus primarily on physical needs rather than broader human or societal development. This mandate presents very real challenges to project sustainability, challenges that in his view and in the view of his agency, ZOA Refugee Care, Netherlands, can only be overcome through meaningful participation by aid-recipients. That said, implementing participatory approaches in areas affected by protracted political conflict is by no means a straightforward task. Further, there remains wide variation in working definitions and understandings of the concept of participation among agency staff and major differences also in its practical application. Ideas about what participation actually entails are often quite imprecise. While most agencies agree that consultation and participation of beneficiaries are key xxiii and there can always be more consultation xxiv, it is not always evident what this amounts to in practice. Within some agencies there is disagreement about the value of and mechanisms for participation. This applies even in the more prominent agencies that have a clear institutional commitment to the approach. Indeed, it is clear that some actors regard the notion as quite problematic: 20