UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of resettlement on community recovery : an analysis of post-tsunami Aceh, Indonesia Panjwani, Dilnoor 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2013_fall_panjwani_dilnoor.pdf [ 8.14MB ]
JSON: 24-1.0165622.json
JSON-LD: 24-1.0165622-ld.json
RDF/XML (Pretty): 24-1.0165622-rdf.xml
RDF/JSON: 24-1.0165622-rdf.json
Turtle: 24-1.0165622-turtle.txt
N-Triples: 24-1.0165622-rdf-ntriples.txt
Original Record: 24-1.0165622-source.json
Full Text

Full Text

  THE EFFECTS OF RESETTLEMENT ON COMMUNITY RECOVERY:  AN ANALYSIS OF POST-TSUNAMI ACEH, INDONESIA by DILNOOR PANJWANI B.A.H., Queen?s University (2005) M.Sc., The London School of Economics and Political Science (2007)  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Planning)  THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver) October 2013  ? Dilnoor Panjwani, 2013 ii  Abstract  In a context of constrained time, resources and geographical space, populations displaced by natural disaster often face diverse and/or ad hoc resettlement schemes.  The purpose of this dissertation is to understand factors that can influence successful resettlement several years after a natural disaster so that it may better inform the management and planning of recovery processes. As such, this research asks: ?How do resettlement patterns influence long-term holistic community disaster recovery?? To address this question, this study explores recovery across five communities affected by the 2004 Indian Ocean Tsunami in Aceh, Indonesia.  Using a mixed methods comparative case study design, villages in Banda Aceh and Aceh Besar that represent differences in resettlement process and pattern were targeted. Findings are based on fieldwork across these communities ? Bitai, Gampong Baro, Lampulo, Neheun Compound and Panteriek Compound ? six years after the disaster. Data collection included key informant interviews (i.e., village chief, elders, etc.), key expert interviews (i.e., members of government, NGOs, and academia), focus group discussions (i.e., villagers), direct observations, and secondary data. In the absence of a generally accepted method to measure community disaster recovery, a survey tool was implemented to assess holistic wellbeing outcomes.  This tool is developed by operationalizing a capabilities-based approach through a series of steps that lead to a multi-dimensional recovery index. Results show that differences in overall recovery across the villages are not explained by either resettlement process (participation versus non participation) or pattern (resettlement in previous location versus in new location).  Further qualitative data analysis displays that resettlement success in the five cases is influenced by (1) location, which shapes livelihood, connectivity and safety, and (2) built environment, which shapes sociability, identity and belonging.  Comparisons across cases highlight that these influences impact recovery through a number of mechanisms of importance, such as access to governance structures, availability of gathering places, and social norms and behaviours.  The analysis also describes how mechanisms are mediated by leadership, proximity and community composition.  The findings support a broader understanding of post-disaster processes, including an emphasis on intangible dimensions and a need to approach resettlement using a lens of ?place?.  iii  Preface This dissertation is original, unpublished, and independent work by the author, D. Panjwani. This study was approved by the University of British Columbia Behavioural Research Ethics Board on March 2, 2011 (Certificate # H11-00096).   iv  Table of Contents Abstract .................................................................................................................................................. ii Preface ...................................................................................................................................................iii Table of Contents .................................................................................................................................. iv List of Tables ....................................................................................................................................... vii List of Figures .....................................................................................................................................viii List of Abbreviations ............................................................................................................................ ix Glossary ................................................................................................................................................ xi Acknowledgements .............................................................................................................................. xii Dedication ........................................................................................................................................... xiv CHAPTER ONE: INTRODUCTION .................................................................................................... 1 1.1. Research objectives ..................................................................................................................... 2 1.2. Research questions ...................................................................................................................... 3 1.2.1. Terminology ......................................................................................................................... 4 1.3. Area of study ............................................................................................................................... 6 1.3.1. The 2004 Indian Ocean earthquake and tsunami in Aceh .................................................... 6 1.3.2. Hazard exposure profile ..................................................................................................... 11 1.3.3. Political conflict ................................................................................................................. 12 1.3.4. Identity and community in Aceh ........................................................................................ 14 1.3.5. Shariah law ......................................................................................................................... 14 1.3.6. Human development indicators .......................................................................................... 14 1.3.7. Policy and practice environment ........................................................................................ 20 1.4. Timing of this study .................................................................................................................. 22 1.5. Dissertation outline ................................................................................................................... 22 CHAPTER TWO: LITERATURE REVIEW AND FRAMEWORK ................................................. 24 2.1. Landscape of hazards and disaster research .............................................................................. 24 2.2. Disaster recovery literature ....................................................................................................... 26 2.2.1. Major trends in existing literature ...................................................................................... 27 2.2.2. An opportunity to build back stronger ............................................................................... 30 2.3. Gaps in existing recovery research that this study seeks to address ......................................... 32 2.3.1. Efforts to understand and measure community-scale recovery .......................................... 32 2.3.2. Insights on the role of post-disaster resettlement on recovery and development ............... 34 2.3.3. The role of participation in the recovery process ............................................................... 38 2.3.4. Need for recovery research specific to the context of the developing world ..................... 40 2.4. Linking gaps to theoretical approaches ..................................................................................... 42 2.4.1. Capabilities approach ......................................................................................................... 43 2.4.2. Place ................................................................................................................................... 45 2.5. Conceptual framework .............................................................................................................. 48 2.6. Summary ................................................................................................................................... 50 CHAPTER THREE: RESEARCH DESIGN AND DATA COLLECTION METHODOLOGY ....... 51 3.1. Approach ................................................................................................................................... 51 3.2. The exploratory trip................................................................................................................... 53 3.3. Field preparation ....................................................................................................................... 56 3.4. Fieldwork .................................................................................................................................. 57 3.4.1. Targeted case selection ....................................................................................................... 57 v  3.4.2. Data collection .................................................................................................................... 62 3.4.3. Journal ................................................................................................................................ 71 3.5. Fieldwork challenges and ethical issues ................................................................................... 71 3.5.1. Researcher ? research assistant relationship ...................................................................... 71 3.5.2. Local expectations .............................................................................................................. 72 3.5.3. Researcher ? participant relationship ................................................................................. 73 3.5.4. Benefit to participants ........................................................................................................ 75 3.5.5. Trustworthiness .................................................................................................................. 75 4.1. Transition from recovery to development ................................................................................. 77 4.1.1. Background ........................................................................................................................ 77 4.1.2. The aftermath of international aid ...................................................................................... 79 4.1.3. Alternative approaches to urban economic growth ............................................................ 85 4.1.4. Societal changes ................................................................................................................. 89 4.1.5. Summary ............................................................................................................................ 92 4.2. Description of communities ...................................................................................................... 93 4.2.1. The compound of Neheun .................................................................................................. 93 4.2.2. The compound of Panteriek ............................................................................................. 104 4.2.3. The village of Gampong Baro .......................................................................................... 113 4.2.4. The village of Bitai ........................................................................................................... 121 4.2.5. The village of Lampulo .................................................................................................... 132 4.2.6. Summary of case descriptions .......................................................................................... 141 CHAPTER FIVE: ASSESSING COMMUNITY RECOVERY ....................................................... 146 5.1. Background ............................................................................................................................. 146 5.1.1. Implementation of a capabilities-based approach ............................................................ 146 5.2. Selecting relevant capabilities ................................................................................................. 148 5.3. Matching selected capabilities to indicators ........................................................................... 150 5.3.1. Community scale .............................................................................................................. 150 5.3.2. Indicator selection ............................................................................................................ 151 5.3.3. Indicator data sources ....................................................................................................... 152 5.4. Rating indicators and index scoring ........................................................................................ 155 5.5. Index validation ....................................................................................................................... 158 5.6. Recovery ranking .................................................................................................................... 160 5.7. Correlations across groups of capabilities ............................................................................... 161 5.7.1. Physical vs. intangible ...................................................................................................... 162 5.7.2. Individual vs. community-driven ..................................................................................... 164 5.7.3. Immediate vs. transition ................................................................................................... 165 5.8 Mean scores and summary ....................................................................................................... 167 CHAPTER SIX: ?PLACENESS? ANALYSIS .................................................................................. 170 6.1 Framing the analysis ................................................................................................................ 170 6.2 Location influences .................................................................................................................. 172 6.2.1. Livelihood activities ......................................................................................................... 173 6.2.2. Connectivity to the broader networks .............................................................................. 178 6.2.3. Safety from hazards .......................................................................................................... 182 6.3. Built environment influences .................................................................................................. 185 6.3.1. Sociability ......................................................................................................................... 185 6.3.2. Identity ............................................................................................................................. 196 6.3.3. Belonging ......................................................................................................................... 204 vi  6.4. Summary ................................................................................................................................. 207 CHAPTER SEVEN: CONCLUSIONS ............................................................................................. 211 7.1. Summary of research findings ................................................................................................ 211 7.2. Contributions to existing bodies of literatures ........................................................................ 214 7.2.1. Understandings of community scale disaster recovery .................................................... 214 7.2.2. Place-related factors that influence community scale disaster recovery .......................... 215 7.2.3. Participation and citizen involvement in a post-disaster context ..................................... 216 7.2.4. Relative disaster recovery across cases ............................................................................ 217 7.2.5. Resettlement ..................................................................................................................... 217 7.2.6. From recovery to development......................................................................................... 218 7.3. Policy and practice .................................................................................................................. 219 7.3.1. Transferability .................................................................................................................. 219 7.3.2. Implications for post-disaster resettlement and recovery ................................................. 220 7.4. Limitations and possibilities ................................................................................................... 224 Bibliography ...................................................................................................................................... 229 Appendix One: Semi-structured interview and focus group discussion templates ............................ 242 Appendix Two: Relevance of indicators to community recovery ..................................................... 245 Appendix Three: Detailed breakdown of MDRI scoring ................................................................... 248 Appendix Four: Tsunami simulation maps ........................................................................................ 250  vii  List of Tables Table 1.1. Damage assessment figures of tsunami in Aceh ................................................................... 8 Table 1.2. Overall ranking of jurisdictions in Aceh by development indicator ................................... 19 Table 2.1. Disaster and development studies ....................................................................................... 41 Table 3.1. Case criteria and communities selected .............................................................................. 61 Table 3.2. Data source by variable ....................................................................................................... 62 Table 3.3. Key informants by village and category ............................................................................. 65 Table 3.4. Focus group discussions by case ......................................................................................... 68 Table 3.5. Experts interviewed ............................................................................................................ 71 Table 4.1. Responses to question on whether recovery was complete ................................................ 78 Table 4.2. Responses to question on state of current situation ............................................................ 79 Table 4.3. Neheun Compound housing breakdown ........................................................................... 101 Table 4.4. Resettlement spectrum ...................................................................................................... 142 Table 4.5. Participation ranking of cases ........................................................................................... 143 Table 4.6. Key descriptors of cases.................................................................................................... 145 Table 5.1. Indicator data sources ....................................................................................................... 155 Table 5.2. MDRI with scores ............................................................................................................. 159 Table 5.3. Case criteria and communities selected (ranked) .............................................................. 161 Table 5.4. Capabilities scores: physical and intangible ..................................................................... 163 Table 5.5. Capabilities scores: individual and community ................................................................ 164 Table 5.6. Capabilities scores: immediate and transition ................................................................... 167 Table 5.7. Mean capabilities scores: summary .................................................................................. 168 Table 5.8. Absolute differences in mean capabilities scores ............................................................. 169 Table 7.1. Overall case rankings ........................................................................................................ 213 viii  List of Figures Figure 1.1. Map of Aceh ........................................................................................................................ 7 Figure 1.2. Damage and loss from tsunami in Aceh (Million US$) ...................................................... 8 Figure 1.3. Banda Aceh spatial plan 2005 ........................................................................................... 11 Figure 1.4. Population growth rate in Indonesia .................................................................................. 16 Figure 1.5. Poverty rate by selected provinces in Indonesia: 1999 ? 2009 .......................................... 17 Figure 1.6. Selected provincial HDI in Indonesia, 1996 ? 2008 .......................................................... 17 Figure 1.7. Gender-related development progress in Aceh, 1996 ? 2008............................................ 18 Figure 2.1. Core topics of hazards and disaster research *Topic of this study .................................... 25 Figure 2.2. Conceptual framework ...................................................................................................... 49 Figure 3.1. Research process................................................................................................................ 53 Figure 3.2. Exploratory trip map .......................................................................................................... 54 Figure 3.3. Village cases on map of Banda Aceh and Aceh Besar. ..................................................... 60 Figure 3.4. Village cases on map of Masjid Raya subdistrict of Aceh Besar ...................................... 61 Figure 4.1. Banda Aceh spatial plan 2016. Big orange circle marks the ?South?. .............................. 86 Figure 4.2. Banda Aceh spatial plan with evacuation routes. .............................................................. 91 Figure 4.3. View of other compounds around Neheun Compound ..................................................... 94 Figure 4.4. Uphill view of incline in Neheun Compound .................................................................... 95 Figure 4.5. Neheun Compound. ........................................................................................................... 97 Figure 4.6. Aerial image of Panteriek Compound before (2004) and after (2009) ............................ 105 Figure 4.7. Shop set up outside NGO house in Panteriek Compound ............................................... 110 Figure 4.8. Renovations to NGO house in Panteriek Compound ...................................................... 111 Figure 4.9. Half constructed bridge from mainland to old Gampong Baro location ......................... 115 Figure 4.10. Raft used by villagers of Gampong Baro to access old location ................................... 115 Figure 4.11. View from top of new resettlement of Gampong Baro ................................................. 120 Figure 4.12. New resettlement of Gampong Baro ............................................................................. 120 Figure 4.13. NGO house in Bitai ....................................................................................................... 126 Figure 4.14. Renovation to NGO house in Bitai ................................................................................ 126 Figure 4.15. Remains of a traditional house in Bitai ......................................................................... 130 Figure 4.16. Traditional wooden house in Bitai................................................................................. 130 Figure 4.17. Wetland in Lampulo ...................................................................................................... 133 Figure 4.18. Traditional fish/shrimp pond next to NGO house in Lampulo ...................................... 133 Figure 4.19. Remains of traditional Lampulo house .......................................................................... 138 Figure 4.20. Temporary shelters being used permanently ................................................................. 138 Figure 4.21. An NGO house next to a traditional house that survived the tsunami. .......................... 139 Figure 6.1. Framing the placeness analysis ........................................................................................ 171 Figure 6.2. Gampong Baro notice board ............................................................................................ 182 Figure 6.3. Shaded hut initiative by women in Block F in Neheun Compound ................................ 190 Figure 6.4. Elderly woman sharing photos of the same location over time ....................................... 192 Figure 6.5. Unused marketplace in Neheun Compound .................................................................... 195 Figure 6.6. Lampulo commemorative symbol: boat on a house ........................................................ 207 Figure 6.7. Summary of placeness analysis ....................................................................................... 209 ix  List of Abbreviations ADB   Asian Development Bank AHDR  Aceh Human Development Report BAPPEDA Badan Percencana Pembangunan Daerah [Regional Body for Planning and Development] BAPPENAS Badan Perencanaan Pembangunan Nasional [National Development Planning Board] BKRA  Badan Kesinambungan Rekonstruksi Aceh [Aceh Sustainable Reconstruction Agency]  BPM  Badan Pemberdayaan Masyarakat [Community Empowerment Board] BPS   Badan Pusat Statistik [Central Statistic Agency] BREB  Behavioural Research Ethics Board  BRR Badan Rehabilitasi dan Rekonstruksi [Agency for the Rehabilitation and Reconstruction] DAI   Domestic Assets Index DII   Disaster Impact Index DRI   Disaster Recovery Index DRR   Disaster Risk Reduction EDFF  Economic Development Financing Facility EM-DAT  Emergency Events Database FBA   Forum Bangum Aceh GAM   Gerakan Aceh Merdeka [Free Aceh Movement] GDI   Gender-related Development Index GEM   Gender Empowerment Measure GIS   Geographic Information Systems HDI   Human Development Index HDR   Human Development Report HPI   Human Poverty Index ICAIOS  International Centre for Aceh and Indian Ocean Studies IAIN   Institut Agama Islam Negeri [Institute of Islamic Countries] IDMC  Internal Displacement Monitoring Centre IOM   International Organization for Migration INGO   International Non-Governmental Organization x  MDF   Multi Donor Fund MDRI  Multi-Dimensional Recovery Index MOU   Memorandum of Understanding MSR  Multi-Stakeholder Review NAD   Aceh Province (Nanggroe Aceh Darussalam) NGO   Non-Governmental Organization NRC   National Research Council PKK   Pembinaan Kesejahteraan Keluargo [Empowerment Family Welfare] PU   Public Works Agency RA   Research Assistant SD   Sekolah Dasar [Indonesian Primary School] SMP   Sekolah Menengah Pertama [Indonesian Junior High School] TCPS   Tri-Council Policy Statement TDMRC  Tsunami and Disaster Mitigation Research Centre TEC   Tsunami Evaluation Coalition  UGM   University of Gadjah Mada UNDP   United Nations Development Programme UNHCR United Nations High Commissioner for Refugees UNISDR  United Nations International Strategy for Disaster Reduction UNU   United Nations University USA  United States of America USAID  United States Agency for International Development xi  Glossary Asli  Original Barat  West Baro  New Becak  Pedicab Desa  Village Dinas  Service Gampong Village Imam  Religious leader Korban Sacrifice Mahr   Mandatory requirement for all Muslim marriages Maulid  Celebration of the Birth of the Prophet Masjid  Mosque  xii  Acknowledgements My first debt of gratitude goes to the people of Aceh. Your kindness, generous hospitality and whole-hearted cooperation and support will always be cherished. Your positive outlook of life in the face of adversity and your faith in the Almighty has been a valuable lesson and life experience.   In particular, I would like to thank Rina Meutia for sparking my affection for Aceh during an internship at the World Bank in Washington, DC and my Aceh family for their warmth and care during my exploratory trip.  Thank you to Dr. Saiful Mahdi and the entire team at the International Centre for Aceh and Indian Ocean Studies for guiding me in my exploration and to Lia Djamal and Samsuar Abang for making the fieldwork process not only possible, but full of precious adventures.  I am grateful to the village leaderships and community members in Bitai, Gampong Baro, Lampulo, Neheun and Panteriek for allowing me to experience their communities and for letting me into their hearts and homes.  My gratefulness goes to all of the Indonesian government officials, academics, and NGO staff for their instrumental advice and wisdom.  My gratitude extends to all of my professors and colleagues at the School of Community and Regional Planning (SCARP) at the University of British Columbia. Your patience, encouragement and invaluable support in my academic pursuits are very much appreciated. I would like to thank my supervisor, Dr. Stephanie Chang, for your continuing guidance, lasting encouragement and constant support. Your unstinting time and knowledge in providing mentorship have been priceless. My immense appreciation and thanks to my supervisory committee: Dr. Leonora Angeles, Dr. David Edgington and Dr. Penny Gurstein, for your generous advice, critique of drafts and timely feedback.  Outside of my supervisory committee, my gratitude to Dr. Bobbi Setiwan and Dr. Helen Cruz, for providing insights that were important in shaping my research proposal.  I would also like to recognize the team at the Institute of Environment and Human Security at the United Nations University, where I participated in the Summer Academy program in 2012.  In particular, Dr. Susan Cutter, Dr. Mohammed Hamza, Michelle Leighton and the PhD participants who challenged me to link research to policy and practice.  Thank you to my funders.  Financial resources and support for this study came from the Social Sciences and Humanities Research Council (Doctoral Fellowship 2010 ? 2012), the University of British Columbia (Four Year Fellowship 2009 ? 2012; Faculty of Applied Science Graduate xiii  Award 2012 ? 2013), the School of Community and Regional Planning (Graduate Entry Scholarship 2008 ? 2009; Brahm Weisman Memorial Scholarship 2011), and the Liu Centre for Global Issues (Bottom Billion Fieldwork Fund 2011).  My immense gratitude to my fellow PhD students at SCARP, in particular my PhD sisters: Lily Yumagulova, Sarah Church, Aftab Erfan and Victora Barr, for your robust collegiality and your strong support over the years.  Thank you to all of my friends for your forbearing support: in particular, to my loyal writing partner, Giselle De Grandis, for your consistent source of inspiration and motivation. I owe gratitude to my colleagues and involvements at Focus Humanitarian Assistance and the Global Facility for Disaster Risk Reduction for fueling my passion for disaster risk reduction.   Thank you to my parents, in-laws and siblings for your boundless encouragement and for instilling in me by example, compassion, integrity, and aspiration to serve the most vulnerable.  Finally, my unending gratitude to my husband, Akber Samji.  Thank you for your enduring patience, for supporting me unconditionally, and for consistently inspiring me to strive beyond my wildest dreams. xiv  Dedication To all those who have been displaced by disaster and courageously strive to reconstruct their place in our unpredictable world. And to my late grandmother (Dadima), who departed during my PhD journey.  While she did not have the opportunity or means to get a formal education, it is her inspiration and determination that have made this educational pursuit a reality.        1  CHAPTER ONE: INTRODUCTION Natural disasters are occurring with greater frequency and devastation, causing catastrophic loss, damage and disruption worldwide.  Disaster trends indicate an overall increase in the number of natural disasters reported and in the number of people affected by natural disaster worldwide (EM-DAT, 2010).  While the increase is visible across the globe, statistics on the number of people affected by disaster by level of development point to significantly higher numbers in developing countries and least developed countries when compared to developed countries (UNISDR, 2006). It is speculated that this imbalance is due to a combination of factors of risk and vulnerability within a context of climate change, urbanization, overpopulation, poverty and development.  Moreover, it is predicted that the majority of the most damaging disasters will continue to occur in developing countries and least developed countries, each followed by a long and difficult path of recovery.  Coupled with the increasing rates of natural disaster events is a growing prevalence of populations facing displacement due to these events.  Some recent examples include the 2010 Pakistan floods, with an estimated over five million people displaced, and the 2010 Haiti earthquake, with an estimated over 300,000 people displaced.1  The Global Estimate report (IDMC, 2013) reveals that in the year 2012 alone, 32.4 million people were displaced by disaster such as floods, storms and earthquakes.  The majority of people displaced (98 percent over 2008 ? 2012) were in developing countries (ibid.).  As a consequence, and as a component of their recovery process, displaced populations are subject to diverse resettlement schemes. For example, relocation may be pursued as a necessary strategy by government agencies to provide displaced communities with a safe place to recover when remaining in previous sites is no longer a viable option. In order to understand the long-term impacts of resettlement patterns on community recovery, this dissertation will explore recovery across five communities in Aceh, Indonesia, six to seven years following the 2004 Indian Ocean earthquake and tsunami.  An exploration of these cases will provide insights into relationships                                                    1 The researcher gained first hand insights on the disaster resettlement and recovery following these disasters through separate research projects in Haiti (2010) and Pakistan (2012). (For example: 1) Hill, A., Bevington, J., Davidson, R., Chang, S., Eguchi, R., Adams, B., Brink, S., Panjwani, D., Mills, R., Pyatt, S., Honey, M., Amyx, P., Community-Scale Damage, Disruption, and Early Recovery in the 2010 Haiti Earthquake, Earthquake Spectra. October 2011, 27(S1); 2) Panjwani, D. and Aijazi, O. An Exploration of the Role of Aid Delivery in Recovery and Resilience in Flood- Affected Pakistan. 9th Canadian Risk and Hazards Network (CRHNet) Symposium. Vancouver, BC. October 24 - 26, 2012.)  2  between resettlement patterns and recovery that may better inform future resettlement strategy.  To situate this exploration, information in this Chapter (One) includes the objectives of the research, information on the area of study and the key research questions of this study. 1.1. Research objectives The overarching focus of this research project is to explore factors that influence the long-term recovery of communities after disaster. The primary aim within this exploration is to identify resettlement patterns that can facilitate holistic long-term community disaster recovery in the context of development.  Substantive knowledge presented in this paper marks one step toward reaching this goal.  The role of place attributes (for example, access to livelihoods and attachment to land) is a key component of this exploration.  As such, it is hypothesized that for long-term community disaster recovery to be successful, resettlement must address several core attributes of place. The driving force in conducting this research is a need to gather evidence to investigate this hypothesis and illustrate differences in resettlement patterns. An underlying objective of this research project is to produce evidence-based findings to better inform decision-makers and facilitate effective resettlement and recovery policy direction. This goal is supported by recommendations that have emerged in academic research reviews (for example in National Research Council, 2006), emphasize a need for detailed post-disaster research examining what actually occurs in communities during and after the recovery process.  The same review also states a need to identify enablers that may allow for more rapid and successful recovery outcomes. During what will later be described as an exploratory trip, these needs were further verified at meetings with regional program advisors at various international agencies in Indonesia early in this research. Conversations included expressions of an on-going need for more independently conducted planning research that can be utilized to improve the management of post-disaster processes, including resettlement planning. These conversations also verified that in common practice, disaster recovery and resettlement planning was mostly considered as a technical exercise wherein authorities assumed that displaced victims primarily need access to new physical permanent shelter. However, in reality, and what this study will explore, is that following a disaster, individuals often have also lost the social, cultural, and physical grounding offered by communities and places of inhabitancy (Brown & Perkins, 1992). Furthermore, this study supports the perspective that it is these elements of social and cultural matter  3  that endow a place with its defining essence and identity, and that become the building blocks of community development after a disaster (Madanipour et al., 2001).  The research questions and framework employed in this study supports the argument that in order to enable long-term holistic community recovery within a context of development, it is vital for the networks of survivors to be reconnected, re-established and/or nurtured while building their relationship with place (Campanella, 2006).  Findings will contribute to conceptual understandings of recovery that incorporate these broader perspectives. 1.2. Research questions  In order to address the objectives of this research that have just been described, the central research question that is explored throughout this dissertation is: ?How do resettlement patterns influence long-term holistic community disaster recovery??  In particular, the dissertation will focus on case studies from Aceh, Indonesia, and address the question: ?In the context of post-disaster recovery in Aceh, what are some of the key factors influencing the successful recovery of communities after six years (i.e., 2004 to 2011)??  In addressing these questions, this study does not try to explain recovery in Aceh, but rather, explores relationships between some of the key factors that surface as influencing recovery in the five communities investigated.   This central research question is broken down in four sub-questions specific to the case studies. The first and second sub-questions will be addressed in Chapter Five and the third and fourth sub-questions will be addressed in Chapter Six.   1. How can long-term holistic community disaster recovery be systematically assessed across cases? 2. What are some of the key factors that influence long-term holistic community disaster recovery at the village level?2  3. How do resettlement patterns enable or hinder development?  4. How are attributes of place addressed in the various resettlement patterns?                                                     2 The investigation into factors influencing recovery focuses on factors at the village level.  External factors (for example, broader political, cultural and economic factors), although not emphasized in this study, are also important.  4  1.2.1. Terminology  The overarching theme pervasive in disaster literature is that of recovery as the return of a post-disaster situation to some level of normalcy or functionality. In leading disaster sociologist Quarantelli?s (1999) literature review on uses of the term ?recovery?, he finds that the word is most often used to imply an attempt to bring or the bringing of a post-disaster situation to some level of acceptability, which is not necessarily the same as the pre-disaster level.  In this study, recovery will imply the return of a disaster situation to a level of acceptability that exceeds living conditions prior to the disaster.  As such, the operational working definition of the concept of holistic community disaster recovery will be the restoration and improvement of living conditions of disaster-affected communities.  Assessments of recovery will be understood within a framework of disaster risk reduction and human development (i.e., the capabilities approach described later). The definition of holistic relates to the notion that things should be understood as a whole and not just as individual parts.  Therefore, recovery will be understood as a complete integrated system (i.e. as a whole), rather than through one particular dimension (i.e. only physical, economic, social, and so on).    For purposes of the study, community will be approached simply as a group of interacting people who are living in a common geographical location. Two of the cases explored in this study will be 'new' communities (i.e., established anew after the disaster) and three will be 'old' communities (i.e., those already existing from prior to the disaster). However, even in the latter, discussions will distinguish between 'original' community members and the 'newcomers' to the community (i.e. displaced survivors) after the disaster.  When operationalized by case studies in the region of research, the community scale will correspond to the Village or Village Compound3.                                                     3 The terminology used in this dissertation uses ?village? and ?compound?.  This terminology is used simply to define the community/case study.  However, in reality there are unique political and historical complexities associated with village classification.  As described in Mahdi (2009), the smallest unit of a community in Aceh was called a gampong, until Soharto?s regime during which community structures were changed to desa.  The term desa is a Javanese term for village structure, which was standardized across Indonesia during this time.  Some villages have become desa while others remain a gampong.  As Mahdi (2009) describes, the difference between the two is that a desa can administer itself and its leader, while a gampong has less sovereignty.  However, a gampong is not recognized in the administration as it is considered the same as desa.  Therefore, for purposes of this study, the disctinction between the two will not be made.  In addition, some of the case went by both desa and gampong.  A?compound? is used to describe a community unit that is distinct from the larger village, and is not technically a gampong or desa, but was refered to as both interchangeably during interviews.    5  However, complexities associated with defining post-disaster 'community' in this way will surface throughout the exploration.   For example, as Christoplos (2006) writes on the post-tsunami context:  Many of the ?communities? formed through resettlement from different locations do not function as communities ? they are merely people living near each other, often with great tensions and distrust? some new communities are being created by resettling people from different devastated areas in a single site (Christoplos, 2006: 76).  Disaster will be defined in accordance with the UNISDR definition of disaster as:  a serious disruption of the functioning of a community or a society involving widespread human, material, economic or environmental losses and impacts, which exceeds the ability of the affected community or society to cope using its own resources (UNISDR, 2010).    The underlying value that this paper will recognize is that natural disasters are in fact not really natural, but rather, a product of relationships between the natural, built and social environment.  Resettlement is most often used by international aid agencies to describe forced relocation to a new area (for example, UNHCR 2004; ADB, 1998).  In this dissertation, however, resettlement will follow Webster?s dictionary?s definition of resettlement, which includes: 1) to begin to live in a new area after leaving an old one; or 2) to begin to use (an area) again as a place to live.  Therefore, resettlement will include a return to a new location or a return to previous location. In addition, resettlement in this dissertation will imply that resettlement marks a permanent location after temporary displacement.  Attributes of place making are the broad number of elements that can be embodied in the practice of (re)creating place. They include for example, access to livelihoods, attachment, and community engagement.  This paper will use the terms 'developing' and 'least developed?, due to common usage by one of the key target audiences of this research, the international donor and aid community. For example, the United Nations Human Development Report classifies Indonesia as a ?developing? country and the World Bank classifies Indonesia as a ?developing? economy.  It is important to recognize that ideological terminology, including developed and/or developing countries remain problematic in nature for a variety of reasons including an inherent implication of inferiority.  However, these discussions are beyond the scope of this paper.   6  1.3. Area of study 1.3.1. The 2004 Indian Ocean earthquake and tsunami in Aceh  On December 26, 2004, a magnitude 9.1 earthquake occurred 30 km below sea level off the west coast of Sumatra.  The earthquake was the result of a convergence between two tectonic plates: the Indo-Australian plate and the Eurasian plate.  The two plates ruptured along a 1,300 km length of what is called the Sunda megathrust, causing the ocean floor to lift and drop and generating a series of massive waves.   The resulting tsunami directly affected 11 countries, with reports of anywhere from 225,000 to 300,000 deaths and 1 million people displaced.     The closest landmass to the epicentre was the province of Nanggroe Aceh Darussalam (Aceh), situated 250 km away and at the northernmost tip of Sumatra, Indonesia, and most severely impacted (see Figure 1.1).  The tsunami reached Banda Aceh, the capital of Aceh, within 45 minutes of the earthquake, causing immediate death and destruction and flattening the entire coast of Aceh within minutes.  Estimates point to at least 124,000 lives lost, 100,000 people declared missing and 550,000 people homeless (IOM, 2005).  Tsunami mortality figures in Aceh show significant disparity between males and females, with three times as many females killed than males (Rofi et al., 2006).    7   Figure 1.1. Map of Aceh Source: After [Zolman, 2011]. Scale unknown Total damage and loss by the tsunami in Aceh was estimated at US$4.8 billion, equivalent to around 80 percent of Aceh's gross domestic product.4  As indicated by Figure 1.2, 78 percent of the total damage and losses were absorbed by the private sector, while 22 percent was borne by the public sector (BAPPENAS and International Community, 2005). The incredible scale of damage caused by the disaster in Aceh is displayed by figures collected in a multi-agency assessment conducted by the International Organization for Migration (IOM) in 2005 and presented in Table 1.1 below.                                                      4 The damage and loss assessment estimated total costs to replace damage and losses caused by the disaster, in other words, the replacement value (Bappenas and International Community, 2005).   Area of study  Epicentre   8   Figure 1.2. Damage and loss from tsunami in Aceh (Million US$) Source: Based on figures in BRR and International Partners, 2005: 184  Table 1.1. Damage assessment figures of tsunami in Aceh Source: Based on figures in IOM, 2005: 1  - 20 40 60 80 100 120 140 160 180EducationHealthCommunity, cultureand religionHousingTransportationCommunicationsEnergyWater and sanitationFlood control andirrigation worksAgricultureand livestockFisheriesEnterpriseEnvironmentGovernanceand AdministrationBank and FinanceSocial Sector Infrastructure and housing ProductivesectorsCross sectoralInitial Destroyed Major damageSettlement area (ha) 173,673       N/A 60,438              Houses (units) 116,880       66,622       14,026              Health faciliti s (units) 693             457           43                    Scho l bu l ings (units) 1,662          765           183                  Religious buildings (units) 2,580          1,109         348                  Government buildings (units) 1,412          997           10,166              M rkets/kiosks (units) 1,416          1,062         109                  Arterial roads (km) 654             180           298                  Neighborhood roads (km) 1,361          459           286                  Provincial highways (km) 603             229           84                    Bridges (units) 2,267          1,508         408                  Aqueducts (units) 9,122          7,571         547                   9  The 2004 tsunami prompted remarkably high levels of international attention, both in scale of response and level of funding (Telford & Cosgrave, 2007).  An estimated US$7.7 billion was committed toward reconstruction activities (Masyrafah and McKeon, 2008).  NGOs flocked to Aceh soon after the disaster and an estimated 463 agencies were involved with implementing post-disaster projects.  These implementing agencies included the Government of Indonesia, 435 NGOs and 27 donors (including UN agencies); of the 435 NGOs that operated in Aceh, 326 agencies (75%) were international organizations (Masyrafah and McKeon, 2008).   For the first three months after the tsunami, reconstruction was coordinated through the Indonesian government?s National Coordinating Agency for Natural Disaster and Refugee Relief in Jakarta; however, it was inadequately prepared to coordinate the relief operation and vast numbers of international agencies involved (da Silva, 2010).  Therefore, the National Planning Agency (Badan Perencanaan Pembangunan Nasional, BAPPENAS) put into place a Master Plan for the Rehabilitation and Reconstruction of Aceh and Nias on April 15, 2005.  One day later, on April 16, 2005, the President of Indonesia, Susilo Bambang Yudhoyona established the Agency for the Rehabilitation and Reconstruction (Badan Rehabilitasi dan Rekonstruksi, BRR) who would be responsible for coordinating and implementing the Master Plan, described as a community-driven recovery program for Aceh.  With a mandate to operate for a four-year period, BRR oversaw the implementation of close to 20,000 projects by hundreds of organizations.  As described later in the dissertation, the BRR ended its term on April 16, 2009.     All of the implementing agencies (i.e., NGOs and donors) worked under the coordination of BRR, and many houses were built by BRR (often in cases when a community member was not granted a house from the NGO that was engaged in building in their village).  Therefore, reference to the BRR and its role will occur at many instances within this dissertation.  The official mandate of BRR was:  To design policies, strategies and action plans, within an atmosphere of transparency and accountability, and to implement them through effective leadership and coordination of the combined domestic and international effort to rebuild Aceh and Nias back better and safer (Subekti, 2009: xi).  Shelter policy came in place six months after the disaster, causing confusion on the provision of assistance.  A key development included an announcement by BRR in June 2005 that families should be encouraged to return to their own land or voluntarily resettle on land purchased by communities or by BRR.  Additional policies were developed as needs and gaps became evident.   10  For example, BRR issued policy guidelines of land acquisition and tenure provision, emphasizing the role of the government in acquiring land for victims who had lost their land.    This study focuses on case studies in the City of Banda Aceh and the Regency of Aceh Besar, both severely impacted by the tsunami5.  Figure 1.3 below indicates a detailed spatial planning map of the city of Banda Aceh (the area of study) developed by the Regional Body for Planning and Development (Badan Percencana Pembangunan Daerah, BAPPEDA), which clearly outlines a ?green zone? ? land that is not to be rebuilt upon due to vulnerability to flooding and future disaster.  Though developed soon after the disaster, the map was never approved or implemented. This was cited by government officials to be largely due to the reluctance of village members to leave the ?green zone? and also due to the lack of enforcement of boundaries during the rebuilding process that had already started (i.e., NGOs were already building houses in the 'green zone'). Reference to the spatial plan will also appear at several instances within this dissertation.                                                     5 The City of Banda Aceh is composed of 9 subdistricts and 90 villages; the Regency of Aceh Besar neighbours the City of Banda Aceh to the South and East and is composed of 23 districts and 604 villages (BPS, 2010).   11   Figure 1.3. Banda Aceh spatial plan 2005 Source: Map provided by BAPPEDA Banda Aceh during fieldwork 1.3.2. Hazard exposure profile Some claim that, ?the tsunami event was a rare occurrence, and hopefully one that will not occur again in the lifetime of those alive today? ? (UNDP, 2010, p. 107).  Whether or not this is true, Aceh remains an area that is prone to natural disasters on an on-going basis and a reality that plays a strong role in the human development context of Aceh.  Situated on the ?Pacific Ring of Fire?, multiple hazard exposure across Aceh is high all year round, with large populations physically occupying hazardous zones in order to be close to their livelihoods.  With a multi-hazard profile, Aceh has suffered from and will continue to suffer from many disasters of smaller magnitude  12  including earthquakes, landslides and flooding that closely impact development initiatives and income-generating activities.  Concerted efforts have been made to develop records of hazard events since the 2004 tsunami event, namely through the Tsunami and Disaster Mitigation Research Centre (TDMRC) located in Aceh.  An analysis of data from one particular database at the TDMRC shows close to 600 disaster events recorded over the time span of 2008 to 2010, including drought, earthquakes, fires, floods, forest fires, landslides, strong wind events, and surge events.  Combined reported victims of these events include 44 deaths and 404 injuries while losses include 2,380 houses destroyed and 5,713 houses damaged (Setyawati, 2010). These figures point to high hazard frequency, exposure and loss. The data supports a need to understand long-term disaster recovery in Aceh within a framework of disaster risk reduction.  1.3.3. Political conflict  Aceh has a deep-rooted history of political conflict that has implications for understanding recovery in the region.  This includes a long history of strong resistance to control by outsiders since the first Dutch invasion of the independent Sultanate of Aceh in 1873. The history of the recent rebellion dates back to 1976 when Hasan di Tiro founded the Gerakan Aceh Merdeka (Free Aceh Movement, GAM) and proclaimed Aceh to be an independent state.  This was followed by 29 years of conflict in Aceh between the Government of Indonesia and the GAM; estimates point to anywhere from 15,000 to 30,000 deaths from this conflict (Reid, 2006). The period was characterized by the presence of Indonesian central government troops, human rights abuses6, and disputes over the allocation of natural resources revenues7; impacts of the conflict included high levels of corruption, weak local government and under-investments in public services8 (da Silva, 2010: 26).  Peace in Aceh has been described as the exception rather than the norm, with the Acehnese having spent 86 of the last 132 years in armed resistance against the Government of Indonesia (Reid, 2006).                                                        6 See next section on identity, and Reid, 2006.  7 The region of Aceh is a source of oil and gas deposits and has extesnisve mining, forestry and plantation agriculture.  For example, the Exxon-Mobil run Arun LNG plant is near North Aceh and is Indonesia?s fourth largest export earner; in 1997, Aceh provided 17 percent of Indonesia?s almost US $12 billion in oil and gas export revenue (Kingsbury, 2006: 6).  8 For further reading on impacts of the conflict, see Aspinall 2006 and Reid, 2006.  13  The official end of the conflict was established on August 15, 2005 by the Helsinki Memorandum of Understanding (MoU) between the Government of Indonesia and the GAM.  The agreement set out arrangements for a number of elements relevant to the governance of Aceh, including political participation, economic management, rule of laws, practice of human rights, and so on (MSR, 2009: 3).  The peace has held to date and through two election cycles. The 2004 Indian Ocean tsunami is consistently cited as helping to trigger peace negotiations between the Government of Indonesia and the GAM for a number of reasons, including the need for cooperation and coordination in post-disaster efforts.  For example, Zeccola (2011) writes:  The tsunami and the conflict entered into a symbiotic relationship on 26 December 2004; the tsunami inspired peace and peace was favourable for tsunami reconstruction.  Both disasters deeply shaped new historical, social, political and economic contours in Aceh (p. 308).   Consequently, recovery efforts following the disaster have been a combination of post-disaster and post-conflict efforts; specific funds were allocated toward projects focused specifically on post-conflict recovery (i.e., through reintegration assistance).  Reintegration assistance has included political reintegration into legitimate political processes and social reintegration through skills training to become constructive citizens (MSR, 2009).  For example, these include projects focused on governance and administration, along with housing and enterprise support (for example, opportunities in the construction sector).  Furthermore, reports show that former combatants and civilian conflict victims have received greater amounts of reintegration assistance than have civilian non-victims (MSR, 2009).   While this dissertation will not go into depth on the complexities associated with lasting peace in Aceh, it will point to several writings that have focused on the political economy of the peace process in Aceh (for example, see Aspinall, 2009; de Alwis & Hedman, 2009; Kingsbury, 2006; Zeccola, 2011). Over the course of the fieldwork, distinguishing between post-tsunami and post-conflict impacts and recovery was difficult; whether or not this is even possible was an ongoing dilemma. To address this challenge, this study describes community recovery from the disaster.  The distinction with post-conflict is not made explicit as recovery outcomes are described as simply wellbeing outcomes.  Repressions of conflict are excluded from the analysis due to insufficient primarily data collected and based on advice received from local scholars over the course of the fieldwork and within the scope of this study.  14  1.3.4. Identity and community in Aceh Identity and community in Aceh has largely been shaped by the political conflict landscape.  Literature on Aceh's place in the Indonesian political arena points to an emphasis on a "distinct Acehnese ethnic identity defined against that of Indonesia? (Aspinall, 2006).  For example, the independence movement based its claims for territorial sovereignty on the construct of an Acehnese identity that had distinctive ethnic, linguistic, cultural, historical and geographically specific characteristics (Miller, 2010).  Extreme violence and abuses by Indonesian government troops during these times of conflict deepened feeling of alienation and strengthened ethno-nationalism in Aceh (Aspinall, 2006).   Figures on displacement as a consequence of the conflict vary.  For example, estimates point to fluctuations from 12,000 and 180,000 people displaced between 1999 and 2003 (Oxfam, 2003).  Major driving forces behind the displacement included the clashes between the GAM, militias, and military forces; some studies provide evidence of economic opportunities as playing an important role as well (Czaika and Kis-Kato, 2009).   The Indian Ocean tsunami and role of the national government in recovery efforts have been accredited for precipitating a stronger sense of belonging to the 'nation' alongside the peace accord; the majority of people displaced by the conflict and disaster have been resettled. 1.3.5. Shariah law One of the key landmarks from the recent GAM conflict includes the passing of Law No.18 in January 2002 that conferred special autonomous status for the province of Aceh Special Region as the province of Nanggroe Aceh Darussalam.  The law sought to convince the Acehnese to remain part of Indonesia by granting the power for self-governance (Bastamam-Ahmad, 2007: 159).  The period marked the implementation of Islamic shariah law, along with greater revenue sharing of Aceh's natural resources (Kenny et al., 2010: 9).  The implementation of Islamic law in Aceh has faced controversy and has resulted in the formation of groups that support or are in opposition to the implementation (for example, see Bastamam-Ahmad, 2007).   1.3.6. Human development indicators A high level understanding of the state of Aceh is provided in an assessment of human development indicators published by the UNDP in Aceh's Human Development Report (AHDR)  15  (UNDP, 2011) 9.  These indicators include Aceh's Human Development Index (HDI)10, Human Poverty Index (HPI)11, Gender Development-related Index (GDI)12 and Gender Empowerment Measure (GEM)13.  While HPI shows a promising picture of development, the three other development indicators (HDI, GDI and GEM) provide a mixed picture of progress and decline. In order to situate this study in the broader context of development, these trends will be discussed in brief, following a short depiction of population growth.   The population of Aceh is around five million people; the population of Banda Aceh is 219,00 people.  While population growth is not necessarily considered an indicator of development, it can point to some interesting trends to provide contextual understanding. As displayed in Figure 1.4, the population growth rate in Aceh has dropped in recent years; the decline has been much faster in Aceh than the rest of Indonesia.  Reasons for the drop included higher mortality, declining birth rate and out-migration; all of these factors can be related to disruption caused by the conflict and a weakening economy (UNDP, 2011: 17).  While some of the villages explored in this study support these trends, they point to complexities associated with population growth at the community scale, including push and pull factors such as access to livelihoods and governance structures.                                                       9 At the request of the Government of Aceh, the Aceh Human Development Report (AHDR) 2010 was prepared by the UNDP with the intention of reviewing progress and serving as a reference for the future implementation of programmes (UNDP, 2011).  The report was developed through consultations with experts, practitioners and local government departments and is a first for a province in Indonesia, though others are planned for the future.  Data presented is up to 2008, the most recent year for which data was available at the time of publication.   10 The HDI was introduced by the UNDP in 1990 as an alternative way of measuring people's wellbeing.  The HDI combines the following dimensions: life expectancy at birth, mean years of schooling, expected years of schooling and Gross National Income (GNI) per capita.   11 The HPI was introduced by the UN in 1997 to complement the HDI.  The HPI for provinces includes the proportions of the population that is not expected to reach 40 years of age, without access to clean water, without access to health facilities, and the proportion of under-nourished children less than five years of age; the HPI for districts also includes level of adult literacy (UNDP, 2011: 5).   12 The GDI was introduced by the UNDP in 1995 to provide a gender dimension to the HDI.  The GDI compares data for men and women on the basis of: life expectancy, adult literacy, years of schooling and contributions to household income.   13 The GEM was also introduced by the UNDP in 1995 to provide a gender dimension to the HDI.  The GEM for provinces includes the proportions of women in the local parliament or assembly, in senior official, managerial and technical staff positions, and in the local labour force; the GEM for districts in Aceh also includes the average non-agricultural wage (UNDP, 2011: 5).   16   Figure 1.4. Population growth rate in Indonesia Source: UNDP, 2011: 16 Figure 1.5 shows that poverty as measured by the HPI in Aceh has steadily declined at the provincial level and among all of the districts from 1999 to 2008.  In addition, an analysis of the poverty rate displays that Aceh is somewhat lower compared to the other regions.  The graphs indicate that levels of poverty have been decreasing since the 2004 disaster.  On the other hand, displayed in Figure 1.6, Aceh?s HDI14 shows an increase from 1996 to 2007 that is in line with national levels.  However, compared to other provinces, Aceh fell from being 9th in national rankings in 1996 to 15th in 2002 and then 29th in 2008 with a drop to 67.1 (UNDP, 2011: 14).  The early rise in HDI, followed by a fall in rankings indicates that the other provinces (that were not impacted by the 2004 disaster) are making faster progress (UNDP, 2011: 14).                                                     14 Wisner in Pelling (2003: 46) describes a mathematical study of over 200 possible indicators of disaster risk vulnerability in order to come up with an index for use in its World Vulnerability Report.  He describes the result to be "striking" and that the HDI turns out "to be the best predictor of deaths due to extreme natural events, world-wide, on average over the twenty years 1980 - 1999".    17   Figure 1.5. Poverty rate by selected provinces in Indonesia: 1999 ? 2009 Source: UNDP, 2011: 15  Figure 1.6. Selected provincial HDI in Indonesia, 1996 ? 2008 Source: UNDP, 2011: 13 As depicted in Figure 1.7 below, the two indicators that are related to gender development in Aceh display some disappointing trends over the period 1996 to 2008.  The GDI shows little progress and remains around the 60 mark and is a few points lower in 2008 than it was in 1996.  The GEM has moved up and down, starting at a high of 57.3 in 1996 but standing at 50.2 by 2008 (UNDP, 2011: 14).  Plausible explanations that are discussed in the AHDR for the mixed picture of gender include numerous advances and retreats in the position of women in Acehnese society. Some advances  18  include the increased capacity of civil society on gender justice issues, greater collaborations with the government on these issues, and the implementation of Shariah courts leading to greater rights for women in respect to inheritance and property.  Some retreats in the GEM index are associated with the return of former male combatants to households, decreasing women's participation in the workplace and the average wage rates for females cited as being lower than males in most jurisdictions.  While most of these explanations are situated in the macro scale (i.e., Aceh province), this study will incorporate gender as a crosscutting theme and explore implications at the micro scale (i.e., the local community).  In particular, it will look at impacts on gender as a consequence of resettlement pattern.  Figure 1.7. Gender-related development progress in Aceh, 1996 ? 2008 Source: UNDP, 2011: 14 As indicated earlier in Figure 1.1, the village cases explored in this study are located in the region of Kota Banda Aceh (City of Banda Aceh) and in the region of Aceh Besar, which surrounds the provincial capital of Banda Aceh. 15   The 2009 populations of these regions are approximately 212,241 and 312,762, respectively (BPS, 2010). As displayed in Table 1.2, both Banda Aceh and Aceh Besar are amongst the top five ranked regions of Aceh based on 2008 development indicators.  Therefore, while findings of this study are relevant for other regions of Aceh impacted by the 2004                                                    15 Locations of the villages will be indicated in later maps.  19  tsunami, they are representative of two regions that are in the higher end of development in the province, even prior to the disaster event.     Table 1.2. Overall ranking of jurisdictions in Aceh by development indicator Source: UNDP, 2011: 21 The figures/tables above and associated descriptions display the development situation in Aceh using standardized measurements and indicators.  However, as witnessed by the author during participation at consultation meetings on the AHDR, the reliability of data that has been used to  20  develop this illustration is questionable and a challenge that was raised by many key stakeholders.  This is due to lack of standardized data collection mechanisms, gaps in data and possible errors in processing data.  The report argues that "while specific numbers may be open to question, the larger trends implied by the data are more robust and are ultimately what is more important" (UNDP, 2011: 5).  While the trends we have discussed do paint a broad picture of the state of development, they do not necessarily capture ground realities from the perspective of community recovery. This study aims to provide a more localized understanding of development by adapting some of the indicators used across the measures that have just been discussed to the community scale.  Findings at the community scale, including in-depth perspectives of communities, can provide insights on reasons for the direction of these broader development trends.  1.3.7. Policy and practice environment  With the sheer magnitude of the destruction and loss of life in conjunction with pressure from the international community armed with extensive funding, government agencies, NGOs and donors were tasked with the recovery of Aceh across sectors in which many did not have expertise, including resettlement activities.16 A variety of strategies were put in place, which resulted in inconsistent processes being followed. Survivors who were previously renters and squatters (i.e., where they were not land owners) were among the poorest and most vulnerable victims of the tsunami, yet, were not included in resettlement policy until June 2006, when BRR recognized their needs (Fitzpatrick, 2007).  It was at this time that BRR introduced a policy of cash grants for renters and squatters.  It was only in 2007 that a policy of land and housing for renters and squatters was announced, enabling them to the provision of land and shelter.  As a consequence, many of them were the last in line for housing assistance, adding complexities to some of the cases that will be explored in this study.   In instances of relocation in Aceh, there was an absence of any sort of formal standardized processes established by the local government or BRR for assessing the suitability of a site that had been identified for development as a relocation site, nor was there a procedure for determining which infrastructures would be put in place (da Silva, 2010: 51).  As described by da Silva (2010: 51), rapid                                                    16 Some examples of NGOs and donors involved in housing and shelter included: Oxfam, Islamic Relief, Turkish Red Cross, Plan International, BRR, Quatar Government, Saudi Government, Chinese Government, UNHabitat, Aceh Relief, Budda Tzu Chi Foundation, IOM, and so on.  21  assessments that were carried out used simple checklists on potential issues including site boundaries, susceptibility to landslide or flooding, ability to provide access to water, and so on.   Many existing academic and practitioner reports have provided critical accounts of what worked and what did not work in the context of post-tsunami Aceh. Most of these focus on the immediate response or short-term recovery.   For example, the Tsunami Evaluation Coalition (TEC) conducted five independent studies in 2005 that led to the identification of a number of weaknesses in the international response.17  Reports allude to valuable lessons such as: Supply-driven, unsolicited and inappropriate aid?unsuitable housing designs and livelihood solutions? aid has led to inequities, gender and conflict-insensitive programmes, indignities, cultural offence and waste?poor understanding among the media and donor public?brushing aside or misleading authorities, communities and local organizations; inadequate support to host families; dominance of English as a 'lingua franca'; 'misrecognition? of local capacities; applying more demanding conditions to national and local 'partners' than those accepted by international organizations; 'poaching' staff from national and local entities; and poor quality beneficiary participation (Telford and Cosgrave, 2007: 21). The governments and the international community alike have been overly optimistic about rehabilitation to development transitions and have failed to consider the real challenges. This is due to lack of analyses of the markets that affect livelihood opportunities and insufficient acknowledgement of how resettlement must link house construction to the development of communities. (Christoplos, 2006: 80) Long-term assessments of recovery outcomes, however, remain limited. Best practices and lessons learned from the recovery in Aceh have been used toward implementing recovery strategy in other major disasters even though long-term impacts of these practices on recovery and development have not necessarily been monitored prior to adapting them in other contexts.  For instance, factors that influence the success of a relocation project have not been evaluated, since several of these relocations were completed some years after the initial international response.  In spite of this, relocations after other disasters since (i.e., the 2010 floods in Pakistan) have followed similar relocation strategies that are primary focused on the provision of shelter and the movement of houses away from a hazard.  This study aims to develop policy implications that will enable better informed                                                    17 These studies included findings on 1) the impact of the response on national and local capacities (Scheper et al., 2006); 2) needs assessment; 3) coordination of the international humanitarian assistance (Bennett et al., 2006); 4) linking relief, rehabilitation and development (Christoplos, 2006); and funding (Flint and Goyder, 2006).  22  relocation strategies that can facilitate positive recovery outcomes after disaster. In doing so, findings aim to address an expressed need for the long-term evaluation by researchers on positive relocation policy (Imura and Shaw, 2009).  1.4. Timing of this study  At the time of fieldwork (2011), the Government of Indonesia considered post-tsunami recovery activities in Aceh to be relatively complete (Subekti, 2009).  This meant that built and physical infrastructure had been restored, international aid agencies had downscaled their local offices and the majority of non-governmental organizations (NGOs) had left. Donor and government funded projects had wrapped up and the national reconstruction agency (BRR) had been closed down since 2009.  Characteristics of the long-term recovery stage are not well defined. However, for purposes of this study, the state of Aceh at the time of fieldwork will be defined as falling in this stage. As few foreigners and outsiders remained in Aceh (for example, NGOs), it was an ideal time to observe the transition between NGO-led recovery and ongoing development. In addition, enough time had passed since the disaster for individuals to speak on the topic without hesitation, yet memories were still vividly clear. The current stage of recovery was also an appropriate time to conduct fieldwork on resettlement, as the majority of affected people had moved into their permanent residences two or more years ago and had built new patterns to give a level of normality to their daily lives. 1.5. Dissertation outline This chapter provided a background on the Indian Ocean Tsunami and the region of Aceh. It pointed to a number of existing challenges and opportunities that motivate this study.  The chapter outlined the research objective, research setting and major research questions that will be explored in this dissertation through a comparative case study research design.   In order to position this study within broader literatures, Chapter Two will discuss major trends in disaster recovery literature and specific gaps that will be addressed through this research.  The guiding approaches and framework for this study will also be outlined in this chapter.   Research design and detailed methodology will be described in Chapter Three.  This includes elements of a comparative case study approach, case study selection criteria, data collection  23  methodologies, and data analyses. Ethical considerations faced by the researcher over the course of fieldwork are also described.  Chapter Four presents the bulk of the data that was collected during fieldwork.  Contextual data on the overall state of recovery in Aceh is included here, along with detailed descriptions of each case study explored in this research.    Chapter Five applies one component of the research framework described in Chapter Two to understand and determine the level of community recovery across the case studies. The chapter will take readers through a process of developing a multidimensional index in order to systematically measure the dependent variable in this study (i.e., community recovery).  Case ranking techniques are used to draw correlations across different dimensions encompassed by the index.   Chapter Six explores independent and intervening variables that influence community recovery outcomes.  This is done through a 'placeness' analysis developed out of thematic qualitative data analysis and is complemented with case comparisons. Specific influencers and mechanisms are described and discussed. The dissertation concludes with Chapter Seven.  The chapter summarizes findings and situates them within the broader literature.  The chapter draws policy and planning implications based on the study, acknowledges limitations of the research and suggests future work.   24  CHAPTER TWO: LITERATURE REVIEW AND FRAMEWORK This chapter reviews and situates the current study in key literatures of disaster recovery.  It further identifies research gaps that this study seeks to address.  Drawing from the literatures reviewed, a conceptual framework that structures an exploration of the research questions outlined in Chapter One is developed.  2.1. Landscape of hazards and disaster research  Professor Gilbert White (1975) and his colleagues conducted the first major survey of research on hazards and disasters over 25 years ago. Their 1975 publication of an Assessment of Research on Natural Hazards was critical in paving the way for an interdisciplinary approach to the research and management of hazards and disasters; the approach has continued to extend across the social sciences (Mileti, 1999; Tierney et al., 2001).  As described in more recent assessments, the amount of research that is currently available on hazard and disaster related topics has increased exponentially over the last three decades, since the first USA assessment (Tierney et al., 2001; NRC, 2006). One of the most recent surveys carried out by the National Research Council (NRC) points to current research spanning across five mainstream topics of hazards and disaster.  As depicted in Figure 2.1 below, these include: hazard vulnerability, hazard mitigation, disaster preparedness, emergency response, and disaster recovery. Four of these - mitigation, preparedness, response and recovery - are also often considered phases of the traditional disaster cycle.  As indicated, this study makes contributions specifically to this study of recovery.     25   Figure 2.1. Core topics of hazards and disaster research *Topic of this study Source: After [NRC, 2006: 2] The overlapping circles and two-directional arrows outlined by Figure 2.1 point to an ongoing need to study hazards and disaster research through an integrated approach with essential interactions among topics (NRC, 2006: 21). Notions of resilience and adaptation surface across these concepts.  While this dissertation focuses specifically on the topic of disaster recovery, this study situates itself within the broader single overarching framework of Figure 2.1. For example, the definition of recovery that has been outlined defines it as ?the restoration and improvement of living conditions of disaster affected communities? and the approach situates recovery within a framework of disaster risk reduction and development.  Disaster risk reduction and development inherently imply reduced hazard vulnerability and increased disaster preparedness through measures of hazard mitigation and stronger emergency response. To provide a broad overview, the next three paragraphs provide brief insights on these topics before recovery literature is explored in considerable depth in Section 2.2.  The arena of hazard vulnerability includes research in areas of physical vulnerability and social vulnerability.  Physical vulnerability includes the study of threats to physical structures and infrastructures, the natural environment and related economic losses (NRC, 2006).   Social vulnerability refers to underlying social conditions that affect a community?s ability to prepare for and recover from impacts of a hazard (Wisner et al., 2004; Anderson and Woodrow, 1998; Cutter et al., 2003).  Current research on social vulnerability has explored characteristics that are associated  26  with social vulnerability and factors that contribute to social vulnerability.  Some examples include gender, age, race, ethnicity, profession, access to resources, social capital, social networks, beliefs, culture, customs, political power, built environment, urbanization, growth rate, economic viability and infrastructure (Cutter et al., 2003, Heinz Center, 2002, Bankoff, 2004). While this study does not directly address physical or social vulnerability and related literatures, the approach acknowledges that in an ideal world an understanding of hazard vulnerability would inform recovery practice in order to address root causes and minimize susceptibility to future disasters (for example, as shown in Yasui, 2007). Hazard mitigation research includes the development of understandings of structural (e.g., building construction and renovating infrastructures) and non-structural (e.g., land-use practices and regulating residence and development in areas of hazard) activities that reduce the likelihood of disasters occurring or the severity of their impacts when they occur.  Discussions on some of these activities will come up at various stages of the dissertation.  For example, relocating communities and continuing to encourage relocation will be discussed as a land-use strategy of the Government of Aceh. While implications of these strategies will be touched upon, the dissertation will not go into the actual impacts from a perspective of hazard mitigation.    Disaster preparedness research includes areas of action taken prior to when a disaster happens and includes tools and procedures for use when a disaster occurs.  Some examples include: formal disaster plans, the training of first responders, the establishment of public education, and financial protection schemes (NRC, 2006).  Emergency response includes activities related to predictions, warnings, evacuations, protections, damage and needs assessments, search and rescue, restoration of public services, and so on (ibid.).  Preparedness activities (i.e., evacuation signs and escape buildings) and emergency response activities (i.e., disaster drills) will be mentioned across the communities explored in this dissertation.  However, impacts of these activities on community recovery will not be covered.  2.2. Disaster recovery literature Many researchers agree that the disaster recovery phase remains the least researched and most poorly understood of the four phases of the traditional disaster cycle ? mitigation, preparedness, response and recovery described above (Berke et al., 1993; Comerio, 1998; Mileti, 1999; Chang, 2010).  As such, research on disaster recovery remains a relatively recent but growing area of  27  academic inquiry. While largely drawing from interdisciplinary areas of inquiry, the vast majority of disaster recovery studies that have been conducted are in the field of sociology (Quarantelli, 1992; Bolin, 1994). However, academic literature on the management of post-disaster recovery processes in the discipline of planning is increasingly being recognized as a new and important area of research and one to which this study aims to contribute (Olshansky & Chang, 2009).   2.2.1. Major trends in existing literature  A review of existing literatures in disaster studies, planning and sociology finds several major shifts in understandings of the recovery process.  It is intended that the direction of this research project will contribute to advancements in understanding disaster recovery alongside these shifts. As elaborated upon below, these include an understanding of recovery as a dynamic rather than stage-like process and an increased recognition of social dimensions (i.e., social elements and mechanisms) of a holistic recovery process. It was not until the 1970s that early studies on reconstruction and recovery were conducted; nevertheless, the area of research remains in its infancy to date (Bolin, 2006).  These initial works on disaster recovery approached the recovery process through sequential stages.  Notably, one of the first comprehensive studies on recovery is work by Haas et al. (1977). In their study on ?Reconstruction Following Disaster?, the researchers develop insights through examining four major disasters ? the 1906 San Francisco earthquake, the 1964 Alaska earthquake, the 1972 Rapid City flood, and the 1972 Managua Nicaragua earthquake. The exploration shows disaster recovery to extend across four chronological periods, claiming that disaster recovery is ?ordered, knowable and predictable? (p. xxvi).  Each period is associated with certain types of activities, as indicated: 1) the emergency period: occurs during the immediate aftermath when the community is forced to cope with damage and destruction in terms of assets, lives and injuries; 2) the restoration period: involves restoring utility, housing, commercial and industrial structures, if possible, as well as a return to relatively normal functioning of social and economic activities; 3) the replacement reconstruction period: occurs when the capital stock is rebuilt to pre-disaster levels, and social and economic activities return to pre-disaster levels or greater; and  4) the commemorative, betterment and developmental reconstruction period: includes activities to commemorate the disaster, fix damages and work toward future growth and development (Haas et al., 1977: 2-3).   28  Over time, researchers have found Haas? classic model to be too simplistic, compared to what reality portrays (Rubin et al., 1985; Bolin, 1994; Vale & Campanella, 2005; Olshansky & Chang, 2009; Edgington, 2010). For example, in their research looking at fourteen case studies of communities recovering from natural disasters in the USA, Rubin et al. (1985) find that the stages overlap and in many cases occur in simultaneous or illogical sequences.  This and other studies have marked a shift away from approaching recovery in a temporal and stage-like fashion to a continuum of dynamic processes.  The shift has contributed to an interest in exploring elements that influence the recovery process and lead to different recovery outcomes. For instance, Rubin et al. (1985) focus on the importance of three key elements that interact to produce a recovery outcome ? the presence of personal leadership, the ability to act, and the knowledge of what to do. Others, for example, emphasize the importance of the nature of citizen involvement, the role of individual volunteers and local leaders, and the influence of rebuilding plans on property rights (Olshansky, 2005).  Some existing studies looking at recovery in post-tsunami Aceh also point to specific elements influencing community recovery outcome (Thorburn, 2010; Mahdi, 2009; Kenny and Clarke, 2010).   For example, Thorburn (2010) points to key determining factors for successful village recovery including elements of governance ranging across leadership (i.e., the breadth and depth of village leadership), decision-making (i.e., number of village meetings), transparency and accountability (i.e., direct village elections). All of these studies support the growing need to look beyond general models of recovery.   As stated by the National Research Council, to date there is little research available on what actually happens in communities during the recovery process or what communities can do to allow for rapid and successful recovery outcomes (NRC 2006: 167).  Similarly, recent research portrays a need to develop greater insight on not only elements in the recovery process, but also the mechanisms involved in recovery (Vale & Campanella, 2005). For instance, Berke et al. (1993) emphasize that recovery processes and outcomes are shaped by factors of participatory decision-making, horizontal community integration (i.e., social networks across community organizations), and vertical integration (i.e., links between communities and higher levels of government).  Similar to these findings, many of the mechanisms that have emerged from recovery observations have highlighted the important role of social mechanisms and social dimensions and the need for further inquiry on these roles.  29  Concurrently, researchers have started to stress recovery as a social process that is based on the social dynamics of societies (Berke et al., 1993; Mileti, 1999).  Many theorists now recognize that disasters are the product of interaction between natural and social factors (Wisner et al., 2004; Pelling, 2003).  As a consequence, understandings of recovery have shifted towards being more encompassing in nature, with increasing attention being placed on the multi-dimensional attributes of the concept, occurring concurrently.  In this sense, recovery is portrayed as, ?simultaneously the production of physical form, the creation of social, cultural and symbolic resources, and the outcome of a negotiated/facilitative process? (Boano, 2009: 3661).  Attention to social dimensions of recovery includes the recognition that damage to the physical environment is often accompanied by the breakdown of social networks, social cohesion and social capital. The role of concepts such as social capital in disaster recovery is an emerging area of recovery research (Nakagawa & Shaw, 2004; Berke et al., 2008; Aldrich, 2012).  For example, Nakagawa & Shaw?s (2004) study on social capital as the ?missing link? in disaster recovery displays promising results on the potential for social capital as an indicator for successful recovery. The researchers develop a model for social capital from the 1995 Kobe earthquake experience and apply it to four different communities affected by the 2001 Gujarat earthquake. Results show that the community with the highest social capital corresponds to the community with the highest satisfaction rate for the new town planning and the fastest recovery rate. Aldrich?s (2012) exploration of four distinct communities ? Tokyo after the 1923 earthquake, Kobe after the 1995 earthquake, Tamil Nadu after the 2004 tsunami, and New Orleans after the 2004 hurricane ? demonstrates that the presence of strong social networks enable a more coordinated recovery process and the ability to rebuild beneficial community ties.    Coupled with the emphasis on social dimensions of recovery is an increased recognition of the concept of holistic recovery.  The concept of holistic implies recovery that ?can address economic, political and social needs ? not only rebuilding infrastructure and housing, but opening the way for more resilient livelihoods? (Wisner et al., 2004: 354). Wisner (2004: 354) also indicates that in practice, the implementations of recovery in this type of way ?requires reversing (or at least substantially palliating) the dynamic pressures and root causes that have contributed to the disaster in the first place?.  Though few studies have approached recovery in such a broad way, this study will do so in accordance to an emerging ?systems approach? ? one that views regions as systems with linkages and interactions across elements such as sectors, geographic space and networks (Olshansky  30  & Chang, 2009).  The particular emphasis in this study will be on key interrelationships between place-based resettlement typologies, attributes of place and recovery outcomes.  In addressing a long-term framework, this study moves away from understanding recovery as an end state or return to normalcy, but rather explores recovery as a process that can be enabled through surrounding factors and conditions.  2.2.2. An opportunity to build back stronger Several researchers have acknowledged that disasters can open rare windows of opportunity for instituting long-term change (for example Olshansky & Chang, 2009) and altering the course of development.  In his paper on the recovery process, Quarantelli (1999: 3) poses the question, ?is it enough to bring back the past, or is something new or different necessary??. From a practitioner perspective, any recovery activity following a disaster that fails to reduce the population?s exposure to risks is ?merely sowing the seeds for future disasters? (IRP, 2007: 3). In the development community, risk reduction is becoming increasingly recognized as an integral component of successful disaster recovery policy and programming. For instance, the UNISDR definition of recovery that has been adopted by many actors in the NGO community is: ?the restoration, and improvement where appropriate, of facilities, livelihoods and living conditions of disaster-affected communities, including efforts to reduce disaster risk factors? (UNISDR, 2010).  The recovery stage provides opportunities for mitigation and perhaps the chance to break the cycle of destruction due to disaster. For example, many researchers emphasize that future vulnerability can be reduced and community resilience improved through incorporating hazard mitigation measures such as developing minimum building codes and land use regulations (Berke et al., 1993; Reddy, 2000). Yasui?s (2007) study of recovery in two communities after the Kobe earthquake demonstrates that certain development practices and capacity building efforts employed during a recovery process reduced overall community vulnerability.  Other researchers insist that a good recovery program begins with a serious commitment to incorporate mitigation and preparedness strategies to reduce future damage (Comerio, 1998: 27). In his study of long-term recovery of three communities of South Carolina struck by Hurricane Hugo in 1989, Reddy (2000) emphasizes that for the successful incorporation of mitigation during recovery, there is a need to adapt mitigation strategies to dynamic local conditions.  This would take into account strategies that reflect the local area, take into account all stakeholder values, and can be adapted to changing political, economic and technological conditions.   31  In reality, the recovery stage brings about unique challenges for both researchers and practitioners, due to compressed time, higher stakes, increased financial resources and intensified public scrutiny (Olshansky & Chang, 2009: 200). Usually, time is the most important factor influencing recovery decisions, actions, and outcomes (Mileti, 1999: 233).  Therefore, the pressure to restore normalcy in response to victims? needs often compromises or leads to the abandonment of what could be promising community development strategies, including vulnerability reduction measures like those described above. There is a common underlying belief that effective rebuilding depends on the speed of construction, albeit this may not necessarily be true. For example, when faced with significant damage and social disruption to a city, priorities tend to change and new possibilities for the future emerge. As a consequence, ?the desire to quickly return to normal, insofar as possible, runs head on into the desire to insure, if possible, that the catastrophe will never be repeated locally? (Haas et al., 1977: 43).  Similarly some view destruction as an ideal time for ?instant urban renewal?, an opportunity to replace a deteriorating area with new, modern lands and structures (Haas et al., 1977: 49). Similarly, the recovery period is a chance to upgrade the quality of construction to resist future disasters and an opportunity to advance programs (such as urban renewal or traffic management strategies) already in place (Mileti, 1999: 236). Therefore the concept of redevelopment can be embodied in theorizing recovery. Redevelopment can be defined as a process that ?alters the design of structures and facilities or the types or patterns of land uses so as to enhance the community in one or more ways? (Schwab et al., 1998: 258). Thus, the disaster context can be seen as an opportunity to speed up plans for change or to bring about new changes to a community.  Also important to note is that disasters do not generate change themselves, but rather often tend to intensify or speed up pre-existing patterns. Researchers have shown that disasters have greater impact on those most vulnerable, and also often exacerbate pre-existing vulnerabilities.  On top of this, certain recovery approaches can induce and/or deepen these vulnerabilities.   In their study of post-disaster recovery following the Northridge earthquake in Los Angeles, Bolin and Stanford (1998) show how pre-existing community vulnerability affected community-level recovery.  Through their case studies they demonstrate that when recovery programs do no address root causes of vulnerability, there is a likelihood of failure. In his study of recovery after the 1976 earthquake in northern Italy, Geipel (1982) finds that disasters highlight existing cultural, social and economic  32  conditions that shape the recovery process. He also found a desire by citizens to return to normalcy, which goes against plans by planners who proposed change, pointing to a need for balance.   2.3. Gaps in existing recovery research that this study seeks to address  In addition to extending the growing exploration of the role of social dimensions in the holistic recovery process, this study seeks to contribute to a number of distinct needs evident through a review of disaster recovery literature.   2.3.1. Efforts to understand and measure community-scale recovery To date there is no consensus definition on what recovery means, how it should be measured, or what constitutes ?successful? recovery (Olshansky & Chang, 2009).  This is because researchers studying recovery continue to struggle with a legacy of conceptual and practical measurement difficulties associated with the concept (NRC, 2006).  These include, for example, variations on whether recovery is a return to pre-disaster conditions, or whether it is approached on a continuum of development.  In spite of this and other challenges, there is a notable need for advancement, particularly in respect to recovery measurement.  For example, in surveying both the research community and practitioner community for their research, Artlikatti et al. (2010) point to an expressed need for a broad-based toolbox of systematic cross-cultural measures of disaster recovery.   Methodologies for measuring recovery across different scales remain limited; existing methods include the use of statistical data (for example, Chang, 2010; Finch et al., 2010; Beniya, 2007), survey tools (for example, Tatsuki, 2007; Bourque et al., 2002); qualitative research methods (for example, Bolin & Stanford, 1991; Ganapati & Ganapati, 2009), computer modelling (for example, Miles & Chang, 2006) and remote-sensing imagery (for example, Brown et al., 2008). The need for qualitative methods are of growing importance out of recognition that though quantitative indicators can show measurable change, it is necessary to have non-quantifiable insight on underlying factors, reasons and processes (Chang, 2010: 322).  As discussed in the previous section, the Haas et al. (1977) model describes recovery to proceed in a linear and orderly fashion to an inevitable end result of recovery.  However,  many researchers since have suggested contrary opinions. For example, in respect to measurement, Rubin et al. (1985: 14) emphasize recovery as an ongoing process and, consequently, something that is difficult to measure one time and have that suffice. As a result and due to recent conceptualizations  33  accepting recovery as a dynamic and endless process, the identification of outcomes and indicators of recovery has become a challenging endeavour (Brown et al., 2008).  Recovery often occurs unevenly across sectors in the community. For example, the local government administration may recover before the local economy, some individuals before others, and so on.  This poses methodological challenges when measuring recovering, especially when looking to measure holistic recovery across a community.   Measurements of successful recovery will face evaluator bias, in which a community might consider their recovery unsuccessful, while the province or higher levels of government deem it to be so, on the contrary (Brown et al., 2008). The degree to which recovery has taken place is very much a matter of perspective and social position.  On the basis of objective social or economic indicators, a community may be considered recovered, whereas in reality social units may not be doing well in objective or subjective terms (NRC, 2006: 149). Brown et al. (2008) caution against prematurely determining the success or failure of recovery, suggesting that it takes several years for observations to be valid. Particularly in the developing world, the recovery process is likely to be strongly linked to funding strategies and timelines of donors and government officials.  Additionally, a key central issue in post disaster recovery is the tension between speed and deliberation with conflicting demands on time (Olshansky & Chang, 2009: 207).  Some suggest that under ideal circumstances all scales be analyzed together such that for example, a study of individual households would include information about their community and a study of community would include an understanding of individual household attributes (Brown et al., 2008: 2).  However, most disaster recovery studies have focused on individual, family and household recovery (Bolin & Stanford, 1998; Bates & Peacock, 1993; Quarantelli, 1999; Tatsuki, 2007).  Consequently, indicators for these scales are more developed than the community scale, with both objective and subjective measures.  For example, Bates and Peacock?s (1993) domestic assets index (DAI) has been used to record changes in a household?s physical assets by measuring changes in available assets used for household activities before and after a disaster (i.e., items within ten functional areas: shelter, food preservation, food preparation, sleeping, human waste disposal, bathing, clothes washing, dish and utensil washing, water heating, and communications).  Evidently very specific to households, there remains a need for approaches to systematically and holistically measure community-scale recovery. This study will contribute to understandings of some of these complexities associated with disaster recovery measurement specific to a community scale.    34  Research on community-scale recovery is more limited (Rubin et al., 1985; Yasui, 2007; Alesch et al., 2009; Chang 2010). In an early study, Rubin et al. (1985) assert community recovery to encompass restoration in all domains of community life: residential, business, public services and facilities, general population, and mitigation.  Mileti (1999) stresses the many dimensions of community recovery to include residential, commercial, industrial, social, and lifelines. Recent work on community recovery supports a systems approach to the concept.  For example, considered a complex component of community recovery is the restoration of infrastructure and lifelines that support system operations that are necessary for normal activities, such as systems that are dependent on transportation, electricity, water, waste disposal, etc. (Peacock et al., 2006). Though frameworks such as those mentioned above provide valuable insights on the nature of community recovery, the question remains as to why some communities recover more successfully than others. This study will explore answers to this complex question. In addition and similar to inquiry on community-scale recovery, there is need for further research on long-term recovery18, given the limited number of studies that have focused on this time frame (Chandrasekhar, 2010; Edgington, 2010).  The time frame is of significant importance in the context of this study in the developing world, due to overlap with an ongoing course of betterment.       2.3.2. Insights on the role of post-disaster resettlement on recovery and development Terminology used by disaster researchers for actions taken in the aftermath of a disaster such as reconstruction, restoration, rehabilitation, rebuilding, restitution, resettlement and recovery are frequently used with inconsistency (Quarantelli 1999).  For the most part, distinctions between and associations among concepts are not yet developed in the disaster literature. This research specifically addresses the nature of the relationship between the concepts of resettlement and recovery.   While studies of post-disaster resettlement are sparse, studies on resettlement are not.  Most investigations on resettlement have been conducted by researchers looking at development-induced relocation due to development initiatives, such as dam construction (McDowell 1996; Oliver-Smith 2009). Many of the studies tend to refer to resettlement as corresponding to relocation, though not all do so (for example, Iuchi, 2010). Of the few researchers who have addressed the topic of resettlement                                                    18 In this study, long term will be defined as over six years after the disaster.   35  of populations after disaster, most prominent is the work of anthropologist Anthony Oliver-Smith. Oliver-Smith (1991: 13) argues that the process of resettlement is far ?more complex than is seen in the approach employed by many reconstruction authorities after disaster?.  He discusses features that lead to the success and/or failure of post-disaster resettlement projects, grouped by factors of site, layout, housing and community input. For example, failures include poor site selection due to easier attainment of land and greater accessibility and topography for rapid construction, at the expense of being situated far from resources, employment, previous village and kin. Layout failures may include monotonous and uniformed designs without adequate consideration for cultural norms and spaces.  Housing issues include inferior use of material, construction that is too small for large families, loss of privacy and inappropriateness for domestic activities. Community input is cited as a strong indicator of success or failure, with minimum consultation leading to a lack of recognition of needs. Successful resettlement is linked strongly with community ownership and responsibility.  Oliver-Smith (1991: 14) alludes to the need for further exploration through concrete cases looking at factors that have a positive versus negative impact on resettlement. In addition, he writes that, ?although disasters account for a significant proportion of resettlement projects, disaster research has yet to fulfill its potential for contributing to this growing body of theory?.  Imura and Shaw (2009: 206) find that ?from the academics? viewpoint, relocation in disaster recovery is a recognized negative activity which international organizations do not fully understand?.   To date, there is no well-accepted model for resettlement in the context of post-disaster recovery.  The few studies that have been conducted on post-disaster resettlement point to a need for further exploration on ways to facilitate effective resettlement (Comerio, 1998; Ingram et al., 2006; Boano, 2009).   The classic and central resettlement model cited by researchers in the realm of development-forced displacement remains the Scudder/Colson model (Scudder, 1985).  Scudder and Colson draw on their work with the Kariba Dam project in Zambia to posit that involuntary resettlement occurs in four stages: 1) planning and recruitment (when planning for resettlement and infrastructure starts and settlers are recruited), 2) transition (when displacement begins), 3) economic and social development (when settlers seek social and economic functions), and 4) handing over and incorporation (when newly established social and economic functions are stabilized).  With monetary compensation being the main element driving resettlement, the model has limited relevance to post-disaster applications.  A newer model, Cernea?s theoretical Impoverishment Risks and Reconstruction model (2000) is directed toward improving the living conditions of affected populations after development-induced  36  displacement.  The model focuses on elements of landlessness, joblessness, homelessness, marginalization, food security, loss of access to common property resources, increased morbidity, and community disarticulation as risks for resettlement failure (Cernea & McDowell, 2000).  Elements demonstrate an overlap with attributes of place, though the model is very much focused on risk as opposed to recovery.  With the lack of a model specific to post-disaster recovery, this study aims to contribute to understandings of resettlement as a component of recovery.  Findings may also contribute to narrowing the gap between the currently divided literatures of disasters and development-forced displacement and resettlement (Button, 2009). Relocation and resettlement policy in international organizations are pursued as more of a development policy than a disaster recovery policy (Imura and Shaw, 2009).  When viewed from a relocation perspective, early works by international organizations recognized that relocation of settlements should be a last option in the recovery process (for example, ADB, 1998; UNDRO, 1982). Resettlement sites may lack meaningful livelihood opportunities, or basic infrastructure, among other social considerations.  Most policy papers and guidance notes from international organizations (for example, Jha, 2009) portray resettlement projects as being negative and meeting frequent resistance, mainly because they posit a future disruption of a community and require further adaptation than simply due to the direct impacts of the disaster. Current researchers and practitioners argue that resettlement ought to be avoided or minimized whenever possible.  Consequences of resettlement may even be considered to cause more grief than the disaster itself (Oliver-Smith, 1991: 13). Nevertheless, resettlement is most often pursued as a necessity, due to the high vulnerability of the population to future hazards. In other cases, however, resettlement may be pursued to manage issues such as those of land use, population growth and economic incentives. Following the 2004 tsunami, some estimates indicate that over 30,000 households in Aceh were unable to rebuild homes on the land that they previously occupied primarily due to flooded land or flooding protection measures (Wegelin, 2006: 5). As a consequence, there are cases of large settlement just outside of city centers, in areas where land was more readily available.  Relocation has been shown to perpetuate and create new vulnerabilities for communities (for example in Oliver-Smith, 2009; Ingram et al., 2006). In their research, Ingram et al. (2006) looked at the impact of the massive relocation of affected populations after the 2004 tsunami in Sri Lanka.  Results showed significant negative social, economic and environmental impacts.  These for example included: 1) limited livelihood opportunities, especially among fishermen who were unable to  37  continue fishing due to the location of the new sites; 2) disrupted communities at sites where different communities with different social, economic and cultural backgrounds were brought together; and 3) new sites threatened by both coastal and inland ecosystems.  Similarly, Boano (2009) has studied post-tsunami Sri Lanka, a context where large numbers of victims have been resettled in villages and new urban settlements and imposed to a completely new geography.  She defines the post-tsunami era in Sri Lanka as having two overlapping phases: the aftermath of the disaster and the aftermath of the relocation. Through her fieldwork in villages of affected areas of Sri Lanka, Boano finds that transitional shelters were built using inappropriate materials that lead to hot, uncomfortable and unhealthy conditions, and that several permanent housing structures were poorly designed and constructed due to lack of skills and expertise.  Additionally, she finds that there was no consideration given to livelihood-facilitating resources in the selection of areas of relocation sites.  She points out other reconstruction issues, such as the need for people to provide documents of proof of land ownership before being granted any reconstruction assistance, subjecting those without these documents vulnerable to discrimination and feelings of helplessness.  While, these works show mostly adverse immediate impacts, there remains a need to gain insights into the long-term consequences and impacts of post-disaster resettlement and reconstruction in disaster-affected areas. Badri et al. (2006) focus on the 1990 Manjil earthquake in Iran, conducting a questionnaire survey on relocated households 11 years after the disaster, in a settlement that later became a town. Findings point to dramatic and complex changes with significant negative socio-economic consequences for affected households such as a loss of access to natural resources, competition for jobs resulting in weakened social networks, reduced cooperation, increased unemployment for farmers and increased poverty. The authors conclude with recommended measures needed to reduce long-term negative impacts of resettlement such as the diversification of economic activities. Abe et al. (2012) explore a post-tsunami resettlement village in Sri Lanka and find that in order for decision makers to enable sustainable relocation, there needs to be special attention given to social impacts such as internal and external human relationships (for example, relationships between a new settlement as a host village), and factors such as the tendency for return an original location   38  2.3.3. The role of participation in the recovery process  In recent efforts to explore social dimensions important to disaster recovery, a number of researchers have started to emphasize the role of participation in the recovery process (Ganapati & Ganapati, 2009; Kweit & Kweit, 2004; Berke et al., 2008; Olshansky, 2005). Some of this literature argues that in theory, when participatory approaches and full stakeholder involvement is incorporated into recovery and rebuilding processes, the outcome is a better informed, prepared and therefore more resilient community (Vale & Campanella, 2005). In-depth descriptions from Edgington?s (2010) study of the 1995 earthquake in Kobe display challenges in implementing recovery plans when local citizens are not involved and left disempowered. A common theme emerging from literatures is that organizations, institutions and policymakers would like to accomplish the goal of community participation.  However, how to effectively do this within the specific conditions, restrictions and time-sensitivity of the recovery process remains a challenge (Davidson et al., 2007).  The importance of participation is considered pivotal in many disciplines.  For example, in planning literature, researchers advocate for participation as it empowers people and communities while enhancing self-sufficiency (Friedmann, 1992). Similarly, principles of public participation in decision-making are considered central to planning practice and are analysized and operationalized through various conceptual tools; for example, the ?ladder of participation? developed by Arnstein (1969) and adapted by Choguill (1996) to fit the context of developing countries.  Choguill?s ladder displays the lowest to highest forms of participation as: self-management (lowest), conspiracy, informing, diplomacy, dissimulation, conciliation, partnership, and empowerment (highest).  The ladder displays that participation can include the support, manipulation, rejection and/or neglect of community demands, and that people?s self-determination plays an important role in improving their own conditions (Choguill, 1996).  These and other literatures point to the importance of empowerment as a core element of true participation.     A deeper understanding on how elements of participation come together on the ground remains a pressing area of exploration in disaster recovery.  For example, insights from Ganapati & Ganapati?s (2009) study on participatory planning in housing reconstruction following the 1999 earthquake in Turkey stresses a need for larger numbers of case studies highlighting the suitability and influence of participatory mechanisms on the overall reconstruction process. Observations from Davidson et al.?s (2007) exploration of four case studies of post-disaster housing reconstruction (Columbia, El Salvador, and two in Turkey) point to discrepancies in interpretations of participation  39  with usage varying to include civil debate, communication, consultation, delegation, partnership, self-help, and/or communal meetings. The analysis finds that the participation of users in up-front decision-making during the project design and planning phases (including for example, the ability to make meaningful choices among options offered) leads to positive building process and outcomes.   However, they find that this level of participation is rarely achieved and therefore the capabilities of the users are wasted.  In his paper based on an international workshop on community risk assessments, Pelling (2007) writes of the need to distinguish between exploitative and emancipatory participatory approaches.  He describes exploitative approaches as the dominant practice, and as merely instrumental mechanisms for reducing financial costs or for extending the sustainability of projects by promoting feelings of ownership.  Emancipatory approaches, on the other hand, seek to ?provide space for stakeholders to develop their own self-confidence and skills to challenge prevailing local and wider structures of domination? and are less common (Pelling, 2007: 375). Several documents explore why meaningful participation was difficult to establish in the context of post-tsunami Aceh.  For example, some reasons include: the complexity of the situation meant there was a need for more comprehensive planning methods than the widespread approach used by humanitarian agencies of simply asking for lists of priorities; participation rewarded those with strong narrative skills portraying themselves as victims; the manipulation of participatory planning processes of the aid community by aid recipients; and the difficulty in combining essential regulatory functions with participatory processes within time-sensitivity (Christoplos, 2006: 75). In another document, reasons cited include: programmes being supply-driven and disenfranchising for people to take ownership of recovery, aid actors having little understanding of behavioural drivers of post-disaster decisions by communities; and aid actors allowing people to only participate in micro scale issues and not wider governance decisions (Dercon and Kusumawijaya, 2007: 1).  Other publications on participation in post-tsunami activities in Aceh allude to the importance of participatory approach in theory, but to the complexities associated with the approach when implemented. For example, Ochiai and Shaw?s (2009) study of the World Bank?s recovery project in Aceh find that the participatory housing reconstruction effort contributed to key factors of speed, quality, socio-cultural concern, management, and cost.  They find that the participatory approach is effective in housing reconstruction in disaster recovery so long as the balance between different players and these key factors are considered and adjusted based on the local situation.  Other documents emphasize that in post-tsunami Aceh, the meaning of community participation in  40  reconstruction activities downgraded rapidly.  For example, many NGOs claimed to use participatory approaches but in reality, only did in its lowest form, simply using the compulsory term to secure funds (Vebry, 2007).    2.3.4. Need for recovery research specific to the context of the developing world A global report by the UNDP (2004) titled Reducing Disaster Risk: A Challenge for Development provides an account of the evolution of natural disasters as a development concern.  They claim that the dominant view until the 1970s was that natural disasters were synonymous with natural events, indicating, for example, an earthquake was a disaster (i.e., the magnitude of a disaster was considered a function of the magnitude of the hazard). The international community was largely focused on the response part of the disaster cycle. From the 1970s onwards, they write that there was widespread realization that the same natural hazard had different impacts on different structures, generating an interest in mitigation strategies for physical structures.  Also, from the 1970s onwards, but more profoundly in the 1980s and 1990s, the focus moved from looking largely at physical resistance to community resilience including social and economic vulnerability. Causes of disasters started to shift from purely the natural event to development processes contributing to and/or creating various vulnerabilities. By the end of the 1990s, it was clear that all development activities had the potential to increase or decrease risks. These practitioner shifts in focus parallel some of the academic shifts described earlier.   The current state in both academia and practice recognizes that ?disasters and development are so closely related that one cannot be understood without the other? (Bates & Peacock, 1993: p.18). A review of related literatures shows that there is increasing recognition and acknowledgment of the intimate relationship between disasters and development (Bates & Peacock, 1993; Lewis, 1999; McEntire, 2004; UNDP, 2004;, 2009). As displayed in Table 2.1, this relationship draws from extensive literatures across many related disciplines.  This present research study aims to contribute to several of these literatures included in Table 2.1.     41   Table 2.1. Disaster and development studies Source: After [Collins, 2009: 3] In spite of the recognition of a close relationship between disasters and development, and as evident throughout this literature review itself, the vast majority of academic disaster recovery research is developed-world focused. However, as stressed in the introduction, disaster prevalence is increasing in the developing world supporting a need for academic research that is specific to the unique context and that has implications for policy development.   This study will contribute to this need largely with respect to recovery measurement and recovery conceptualization.  For example, in Academic classification Examples of key topicsBuilding designUrbanisationTransportSocial integration, cohesion and exclusionVulnerability, coping and resilienceGenderCulture and developmentSustainabilityDevelopment paradigms and discourseSpace and locationScale of human interactionsIdentity and representationEmergency response systemsHumanitarian assistanceEmergency communicationsInfectious disease managementSocial and psychosocial careFood and nutrition securityWar, peace and conflict resolutionPolitical economy of disasterRisk governanceBiogeographical processesWater managementLife cycle analysisLandform processesGeo-tectonicsPredictive modellingPhysical GeographyThis studyOtherPhysical PlanningSocial and Behavioural StudiesDevelopment StudiesHuman GeographyEmergency ManagementPublic HealthHistorical and Political StudiesEnvironmental Management 42  studying the 1995 earthquake in Kobe, Chang (2010) uses census and other statistical sources to measure community recovery with respect to population, businesses, economic production, income, and port traffic.  How similar measurements can be made in a context of limited availability of reliable data will be addressed in this study through the assessment approach used in Chapter Five.  Furthermore, links between recovery and development are under-researched. For example, based on observations of various post-disaster contexts, Pantelic (1991) has argued that recovery can ?effectively unify development and recovery goals? through various elements like improving disaster resistance of physical structures, increasing standards of living, generating new jobs and skills, and integrating development and recovery goals into a community?s social and cultural values and resources. Concrete evidence to support these inferences is lacking.  In his book, Lewis (1999) writes on development in disaster prone places. He discusses the ?continuum from relief to development?, with reference to popular literatures (for example Linking relief to development: Disaster response with foresight: OFDA, 1997; Linking relief and development: IDS, 1994). Lewis (1999:134) writes of development being regarded as ?the goal when everything else [related to relief and rehabilitation] has been accomplished? and emphasizes the need to approach development as being integrated with relief and reconstruction. This implies that rather than approaching development as being on a continuum of relief and recovery, the three topics may be better approached as occurring simultaneously. He emphasizes the need for greater research on understanding how development processes could be modified to accommodate disaster responses, to reduce vulnerability and to link development to a reduction in the need for relief.   2.4. Linking gaps to theoretical approaches In order to address the gaps in literature and research needs that have been outlined (i.e. insights on understanding and measuring community-scale recovery, the role of post-disaster resettlement on recovery and development, and recovery in the context of the developing world) this study applies a theoretical framework that is the combination of two distinct yet related approaches.  Firstly, it draws upon a capabilities approach as a means to understanding and measuring holistic disaster recovery within a framework of development.  Secondly, it draws upon a place approach as a means to explore resettlement for all of its separate parts and in a manner that emphasizes the human experience.  Each of these approaches will be described in brief.     43  Built upon a capabilities approach and a place approach, the overarching human development paradigm adopted in this study is about nurturing an environment that enables individuals to live productive lives and develop their own potential through means such as enhancing people?s choices. Critical choices stressed in a human development approach include the ability to live a long and healthy life, the ability to acquire knowledge and the ability to attain a decent standard of living. To put a community spin on human development, for purposes of this research, implies an emphasis on elements that provide a supportive environment of networks and services available to the community that enable choice. 2.4.1. Capabilities approach The first approach adopted in this study is the capabilities approach.  Developed by economist Amartya Sen and philosopher Martha Nussbaum (see Sen, 1999; Nussbaum, 2000), the capabilities approach remains a widely accepted paradigm in development.  The approach has been instrumental in changing the perspective of development among academics and international agencies, by arguing that the standard of living of individuals in a society is determined by "the ability of people to lead the kind of life they have reason to value" (Anand & Sen, 2000).  At its core, the approach supports the idea that standard of living is a measure of the life an individual leads and the amount of choices the individual has for different life achievements (Jasek-Rysdahl, 2010).  As one of its most significant applications, the capabilities approach is the theoretical framework used for the United Nation's Human Development Index (HDI) indicators discussed in Section   Two of the core interrelated concepts in the capability approach include those of functionings and capabilities.  Functionings are used to describe the current life condition of people, while capabilities are the set of possible functionings from which an individual must choose (or, in other words, the freedoms to realize these functionings). For example, eating and starving would be considered functionings, while having the means to obtain an adequate amount of food would be the associated capability.  The capabilities-based approach measures development as a function of the level of wellbeing of a society in terms of functionings, or in other words, the approach observes development as a ?process of expanding the real freedoms that people enjoy? (Sen, 1999: 3).  Sen (1999) conveys the goal of development as expanding capabilities under the following freedoms: 1) economic opportunities, 2) political freedoms, 3) social facilities, 4) transparency guarantees, and 5) protective security; Nussbaum (2000) identifies ten key capabilities from which to  44  approach development that are spread across the following dimensions: life, bodily health, bodily integrity, senses imagination and thought, emotions, practical reason, affiliation, other species, play, and control over one's environment.  This study adapts these dimensions to generate its own list of capabilities for purposes of measuring recovery across the village cases.  The opportunity to apply a capabilities approach to recovery allows for the validation of what matters in terms of recovery from the perspective of a survivor, or in this case, a community of survivors.   How this approach is operationalized in order to assess community recovery outcomes (i.e., to develop community recovery indicators) will be discussed in Chapter Five.  Effectively doing this will have to overcome the widespread challenge expressed by many researchers that, while theoretically attractive, the capabilities approach is difficult to operationalize empirically due to its underspecified nature (Kuklys, 2003; Robeynes, 2006).  As Sen (2005: vii) himself asserts, ?there are widespread doubts about the possibility of making actual empirical use of this richer but more complex procedure?. Implementing this approach for purposes of the present research study also aims to contribute to growing literatures on operationalizing the capabilities approach (for example, Jasek-Rysdahl, 2010). Capabilities approach and disasters To situate community recovery within the ideology of development means not only a return of living conditions to a level of normalcy, but also the process of expanding community capabilities. The idea of conceptually applying capabilities to disaster research is not a new one.  For example, Wisner et al. (2004) applies the concept of entitlements to hazards in his book on vulnerabilities and disasters.  However, only recently have researchers explicitly explored a capabilities-based approach to disaster recovery measurement.  For example, Gardoni & Murphy (2008) use a capabilities-based framework to assess the impact of disaster, track progress in recovery over time, and evaluate the effectiveness of the recovery process.  As such, they use measurements from their capabilities-based Disaster Impact Index (DII), as a benchmark in calculating their Disaster Recovery Index (DRI).  Directly applying their framework to this study will not be feasible given an inability to gather necessary time series DII data.  However, the data collection methodology that will be used in this study to measure community recovery levels across cases will adopt some of the main considerations behind the DRI tool.  These include practical implementation notes on the selection of capabilities and indicators that will be described in Chapter Five.   45  This study recognizes that the capabilities approach is not a theory that is able to explain poverty, inequality and well-being, but rather it is one that can provide concepts and a framework that can help to conceptualize and evaluate these concepts (Robeyns, 2006: 354). Therefore, in this study the capabilities approach will simply be used to develop community recovery indicators in order to enable a multi-dimensional understanding of holistic recovery. Though capabilities described by Sen (1999) and Nussbaum (2000) have been developed for application at an individual scale, the framework will be adapted to a community scale.  Uses at the community scale remain limited, though its close relationship with community based approaches, such as asset mapping and capacity inventory have been researched (Jasek-Rysdahl, 2010).   2.4.2. Place The second approach adopted in this study is that of place.  What separates space from place is that while space is an abstract area, place is a particular part of that area which is endowed with meaning by people (Madanipour, 2001). Disciplines of sociology, geography, social anthropology, planning, environmental psychology, philosophy and architecture each portray a diverse array of perspectives and understanding of place.  Consequently, place literature is multidisciplinary and characterized by tremendous diversity in theoretical, methodological and paradigmatic approaches pointing to usage of the concept in a way that best fits inquiry (Trentelman, 2009).  This study takes the concept of place from various disciplines described below and applies a lens of place to this study of disaster recovery and resettlement.  As a humanist geographer, Tuan (1977: 54) describes place as a calm center of established values and an organized world of meaning (ibid.: 179). In his 1976 publication on Place and Placelessness, Relph argues that "to be human is to live in a world that is filled with significant places; to be human is to have and to know your place.?  His significant contribution to the study of place emphasises the role of peoples' identity of a place in relation to its "persistent sameness and unity which allows that [place] to be differentiated from others" (Relph, 1976: 45).  Political geographer Agnew (1987) outlines the fundamental aspects of place as a meaningful location to include: location (i.e., fixed coordinates), locale (the material setting for social relationships), and sense of place (the subjective and emotional attachment people have to places).  These aspects of place will shape data collected on place attributes.   46  In planning theory, the qualities of places and the qualities of place-making processes are considered ?long-established preoccupations in planning thought? and are heavily valued across the discipline (Healey, 2001: 265). For example, the philosophy of spatial planning advocates for the improvement of quality of life through the promotion, management and regulation of place-making (Healey, 2001).  There is an increased interest in the ?qualities of places?, particularly on the part of policy makers for reasons that are varied and from multiple perspectives. For example, some argue that intense globalization has necessitated a need for a clearer sense of identity, often rooted in particular place; this rootedness in place plays a significant role in the sense of wellbeing of people who are constantly exposed to a multitude of global forces outside of their control (Madanipour, 2001).  In their book on the art and practice of place-making, planners Schneekloth and Shibley (1995) argue that the main goal of place-making is that of building relationships between people and between people and their place.   As an analytic concept for sociology, Gieryn (2000) defines place in relation to three aspects: geographic location, material form and investment with meaning and value.  The demise of ?place-based communities? has become a popular concern in sociological literature since the industrial revolution and the consequential rapid urbanization; in recent decades, economic globalization and advances in communication technologies have exacerbated the demise (Bridger & Alter, 2006).  Similarly, in his book on The Geography of Nowhere (1994: 15), Kunstler argues that the American landscape has become ?a landscape of scary places, the geography of nowhere, that has simply ceased to be a credible human habitat?. The notion of places with no character, no history, and no community appear throughout the book.  This study follows claims that the diverse meanings of place and space are socially constructed by multiple agencies and actors (Madanipour, 2001). In the case of resettlement after disaster, these actors include aid organizations, government agencies and community members.   This study supports the notion that any location may have multiple meanings of place layered over it (Healey, 2001). Therefore, it can be argued that place is embedded in social processes and thus socially constructed. As such, conceptualizations of place link together the social experience of being in a place, the symbolic meaning of qualities in a place and the physical nature of the forms in it (ibid.).  Furthermore, experiences of place are led by notions that ?who we are and how we live emerge in the context of our relation to the places we inhabit and the artefacts with which we surrounds ourselves? (Spelman, 2008: 144). As such, there is a complex relationship between place,  47  architecture, and memory (Steinberg & Shields, 2008). Memory has been described to be ?naturally place-oriented or at least place-supported? (Casey, 1987: 187).  Architecture can act as a metaphor with which to explore place, in that as we move through architecture we experience place and endow places with meaning. The loss of a sense of belonging associated with disaster, both in an emotional as well as a physical sense of loss of physical belongings, act as a barrier to maintaining continuity with one?s past. Just as Cresswell (2004: 11) argues in his Short Introduction to Place that place is ?not just a thing in the world but a way of understanding the world?, a place lens guiding the framework seeks to do the same in understanding disaster recovery. An exploration to understand place, as manifested in the five communities, is examined in this study.  Cresswell goes on to write:  ?place is also a way of seeing, knowing and understanding the world.  When we look at the world as a world of places we see different things.  We see attachments and connections between people and place.  We see worlds of meaning and experience...(Cresswell, 2004:11).    As illustrated in the framework of this research, each stage of this study is shaped by a similar perspective in an attempt to better understand the complexities of resettlement and recovery. By applying an approach of place, this study seeks to understand how place is manifested in resettlement, and what that means for disaster recovery.  In implementing this approach, the focus will be on the location and material settings of communities, social relationships within that setting, and relationship between communities and their location.  As such, the psychosocial dimension of place will not be a strong focus.  However, elements of the humanistic dimension of the concept of place (i.e. emotional bonds) will be conceptually captured within broader dimensions described in Chapter Six. Place and disaster recovery Few researchers and works on disaster recovery have looked at concepts of place in the context of post-disaster initiatives. For example, in a chapter by Zetter and Boano (2010), they examine space and place after natural disasters and forced displacement.   The authors argue that disasters provoke ?placelessness?- the loss of the sense of place - yet that space and place are rarely recreated in cases of forced displacement. Using evidence from post-tsunami reconstruction projects in Sri Lanka, India and Indonesia the researchers demonstrate how concepts of space and place are sacrificed in post-disaster housing developments, representing the use of a narrow and technical view of what housing encompasses and thus, point to opportunities for further inquiry.  Taking these ideas  48  further, this study explores place as it manifests in resettlement patterns and as it relates to community recovery outcomes.  In his chapter on space, place and resilience in post-tsunami Aceh, Mahdi (2009) examines two communities in Banda Aceh during displacement and resettlement in their original villages.  Through comparison, he finds that the main factors that shaped the formation of a new community in a new settlement included social cohesiveness prior to disaster, leadership during the emergency period onward, and interaction with outside intervention in the context of emergency.  He concludes that a "sense of community and interconnectedness among its members is the soul that makes a 'space' a 'place' to live" (Mahdi, 2009: 119).  In her chapter on the remaking of neighbourhood in Banda Aceh after the tsunami, Samuels (2010: 221) demonstrates how social aspects of places, through neighbourhood social relations, activates, and emotional attachments to place are especially important for 'rebuilding everyday life in the reconstructing city".   Cox and Perry (2011) describe a social-psychological process of reorientation to show the critical importance of place as an orienting framework in the disaster recovery of two rural communities affected by a wildfire in Canada.  They display the relevance of place, social capital and identity formation in the psychosocial recovery process of disaster survivors and communities and argue for a reconfiguration of disaster recovery to one that considers the role of place in the recovery process. This study will not delve into psychological meanings of place.  However, the relationship between communities and their surrounding will touch on the surface of emotional elements of place in a way that is more in line with perspectives of geography, planning and sociology mentioned earlier.   2.5. Conceptual framework   Based on the literature review and theoretical approaches that have just been described, the underlying framework that this research will follow is displayed in Figure 2.2 below.  The figure depicts three circles, with the inner ones contained in the outer circle. This illustrates resettlement as an element of recovery, and recovery as an element of development. As such, the framework follows an approach to recovery that embodies elements of improved living conditions.  Situating resettlement within recovery allows for an approach that distinguishes the resettlement process from the larger recovery process while recognizing that resettlement processes condition overall recovery  49  outcomes. Situating recovery within development will contextualize the research questions in a framework of ongoing development.   Figure 2.2. Conceptual framework Source: Author Within a context of development, the assessment of recovery will be carried out through the lens of capabilities.  The figure points to place, in the sense that was described above, as an appropriate lens for resettlement.   This study draws on several elements of this framework at various stages of the research and dissertation.  For example, not only does this study explore resettlement through the lens of place when collecting data, but it will also test the suitability of using a placeness approach for resettlement when analyzing data. Similarly, the appropriateness of using a capabilities approach for measuring community recovery will be examined. The framework will allow for broad understandings on the relationship between place and development through observations gathered on place attributes and development capabilities. As such, linkages between a capabilities approach and a place approach will be explored through the research design. At the same time, the framework will allow for a more detailed multi-dimensional analysis on specific place attributes and capabilities that surface as being most important in resettlement and recovery.    50  2.6. Summary  This chapter reviewed existing literatures and demonstrated a need for greater inquiry on social dimensions of community recovery in the context of the developing world.  Needs identified included: greater understandings on factors that impact community recovery, insights on measuring holistic community recovery, and the role of resettlement on recovery outcomes.  The unique conditions of resettlement and recovery in tsunami-affected regions of Aceh described in Chapter One provide an opportunity to effectively address all of these needs through exploring community recovery in the region.  The next chapter will describe the methodology pursued to carry out this exploration.  A combination of a capabilities approach and a place lens was described as an appropriate way to contribute to filling gaps in literature and framing a study that explores social dimensions of recovery.  Details on how this framework is operationalized in this study appear in later chapters.  For example, Chapter Five will operationalize the capabilities approach through the development of a tool to measure community recovery across the five cases, and Chapter Six will operationalize a place approach through a placeness analysis of the five cases.    51  CHAPTER THREE: RESEARCH DESIGN AND DATA COLLECTION METHODOLOGY This study draws findings through the exploration and comparison of five cases of resettlement in post-tsunami Aceh.  As such, the research employs a comparative case study research design.  Chapter Three provides a detailed account of the research design and methodology that was used to investigate the research questions proposed in Chapter One.  3.1. Approach Given the large-scale devastation and loss of land due to the tsunami, different patterns of resettlement have emerged during the recovery process, presenting an opportunity to draw valuable comparisons. Comparative case studies are an appropriate way to test theory and to develop concepts.  In this study it is used as such, and as a method to explore variations in variables of interest and to allow for comparisons of variables across cases in a systematic way. Comparative case study designs have been used by a number of researchers studying community recovery from natural disasters, for example, work by Bolin (1994) and Rubin (1985).  Its current usage in the field, however, remains relatively limited. Several academic scholars such as those on the National Research Council?s Committee on Disaster Research in the Social Sciences (NRC, 2006) advocate a need for more comparative research in order to document and better understand variations in disaster recovery processes and outcomes.  Others point to the limited capacity to make generalizations on effective policy due to a lack of disaster research that uses multiple case research design with the same methods and variables in each case, thus enabling comparability (Berke et al., 2008).  In addressing this need, this study contributes to the body of research considering uses of comparative approaches in understanding the dynamics of disaster recovery. A case study approach has been chosen for a number of reasons.  It is a preferred strategy when ?how? or ?why? questions are being posed, when the researcher has little control over events (i.e., it is not a controlled environment), and when the focus is on a contemporary phenomenon within some real-life context (Yin, 2003). The approach allows for in-depth inquiry and understanding through multiple data collection methods over a period of time (Stake, 1995). As cases are bounded by time, the approach allows for a compromise between breadth and depth when under time and funding constraints, as is the nature of doctoral research. Case study research allows for  52  applications beyond the specific context of the particular case being studied, as findings can be used to make broader generalizations (Yin, 2003). With the hope of providing contributions to the practice of recovery and resettlement, the approach will enable the development of recommendations based on outcomes.  The case study selection for this study is based on a careful sample of the total number of resettlement villages in Banda Aceh and Aceh Besar through distinct criteria. Details on this will be presented within the section on 'Targeted Case Selection'.  This research follows a mixed methods approach. Both quantitative and qualitative approaches are employed simultaneously to ensure more comprehensive research. Quantitative techniques (such as the use of an index tool), are used to explore relationships among measurable variables in order to explain and predict phenomena, while qualitative techniques (such as observational technique) are used to seek a better understanding of the complex relationships (Leedy & Ormrod, 2001).  As such, data collection has been structured in a way that enables investigating the hypothesis (that for long-term community disaster recovery to be successful, resettlement must address several core attributes of place) through structured observations and by systematically gathering data by variables of recovery, resettlement and placeness. Relationships between variables and underlying reasons have been explored concurrently through in-depth inquiry.  A mixed methods approach has allowed the inquiry to be operationally bound with defined variables, but at the same time has enabled the inquiry of patterns of unanticipated relationships alongside speculated ones (Stake, 1995).  The range of data collection methods within the mixed methods includes conducting fieldwork across village cases (using key informant interviews, focus group discussions and direct observations), key expert interviews in government agencies, academia and NGOs, reading NGO and government reports, and analysing documentation from village offices.  This research is not ethnographic in nature. However, given the amount of time that was spent in the communities of study (approximately one month each), several elements of an ethnographic approach have informed the analytical approach and data collection. For example, a detailed daily journal of observations and experiences was kept throughout the entire research process. These entries allowed for not only ongoing research clarifications and amendments, but also for personal insights on evolution as a researcher over time.  The process of research for this study is portrayed in Figure 3.1.  Subsequent to reviewing existing literatures to identify, justify and situate preliminary research questions, this study consisted of six successive steps: an exploratory trip to the field, the refinement of research questions, field  53  preparation, fieldwork (data collection), data analysis and the dissertation write up. A trip back to the Aceh to share findings is anticipated in the near future (i.e., for the International Centre for Aceh and Indian Ocean Studies biannual meeting in 2015).  While feedback from the trip will not be incorporated into the PhD dissertation, it is expected that they may shape future research initiatives in the region.    Figure 3.1. Research process Source: Author  3.2. The exploratory trip An exploratory trip to Indonesia was conducted over October and November of 2010.  The main goals of the trip were: to gain insight on potential case studies and sites for research design; to build relationships and networks within international and local NGO and donor communities in the field; to firm up institutional affiliations to enable a mutually beneficial research experience; and to familiarize with the culture, language and context in which research was situated. Each of these goals will be addressed below, with reference to research design and fieldwork. Travel in Indonesia included visits to parts of Yogyakarta, Jakarta, Banda Aceh and Aceh Besar (see Figure 3.2).  The majority of the trip was spent in Banda Aceh and Aceh Besar, in regions  54  that had been devastated by the 2004 tsunami. Based on conversation with UNDP staff, local scholars, and NGOs, visits to several surrounding villages of potential interest were arranged.  These visits were mostly observational in nature and provided contextual insights as well as information on the feasibility of research objectives.  Given the widespread destruction and tremendous international community attention, great diversity in resettlement patterns and recovery outcomes in the region were clearly visible.  Some informal group discussions with villagers were conducted in order to grasp a basic understanding of the context and major issues faced by communities.  A list of possible village cases for this study was developed based on these understandings.   Figure 3.2. Exploratory trip map Source: Wikipedia, 2013. Scale unknown While in Yogyakarta, the possibility of conducting comparative research outside of Aceh with cases of village reconstruction in Yogyakarta following the 2006 earthquake was explored. Through participation as an observer in a training workshop on disaster risk reduction, guided field visits to villages were completed.  It became apparent through observations that the recovery process in Yogyakarta was quite different from that in Aceh, with only one case of relocation, and the vast majority of village reconstruction being community-driven with minimal international aid community involvement. As such, it was decided that cases from Yogyakarta would not be used in  55  the research study. Given the vast number of cases from the 2004 tsunami, comparative research would be possible in Aceh itself.  Several elements of the research design required access to persons and documentation across international and local NGO and donor communities. While in Banda Aceh during the exploratory trip, the focus was on creating networks among some of the leading agencies during the recovery process, including UN-Habitat and UNDP. These were all based in Banda Aceh.  Insights from key players from the recovery process at these two agencies led to the validation of some findings in the literature review and pointed to new complexities that would be important in conducting research. For example, the UNDP expressed concern on the high rates of empty homes in many of the reconstructed villages due possibly to a lack of satisfaction with housing design. The UN-Habitat questioned the value of their time-intensive community-based approaches versus the more common contractor-driven approaches used by the majority of NGOs involved with housing reconstruction. While the former enabled greater community engagement during the process, the latter allowed for faster results and therefore was considered perhaps more successful.  There was also curiosity among NGOs with regards to tracing where displaced populations currently resided.  A visit to the local World Bank office in Aceh brought to light the difficulties in the availability of reliable data for academic study. Participation as an observer during a roundtable discussion at UNDP led by government officials on a draft of the Aceh Human Development Report (UNDP, 2011) provided insights into the development context of the current day.  During the trip, institutional affiliation was established with the International Center for Aceh and Indian Ocean Studies in Aceh (ICAIOS) in Banda Aceh.  ICAIOS is a joint effort between three universities in Aceh (Syiah Kuala University, IAIN Ar-Rainry, and University of Malikussaleh), the Indonesia Ministry of Research and Technology, the Government of Aceh, and a number of international academic institutions. Initial contact was made through arranging a face-to-face meeting with the Director of ICAIOS.  The possibility for affiliation was welcomed with enthusiasm and was followed by networking with and insights from others hosted by ICAIOS.  ICAIOS played a supportive role throughout the research project, with technical visa sponsorship, logistical details, PhD working space, discussion forums alongside other international doctoral students, and assistance in facilitating communication with contacts relevant to the research. Housed in the Syiah Kuala University campus, the affiliation gave access to other faculties and researchers in similar and related realms of study. As such, affiliation at ICAIOS provided the  56  opportunity to join a community of international and local researchers studying Aceh. The ICAIOS library was available for use and contained a significant number of writings on Aceh, the military conflict described in Chapter One, the 2004 tsunami, and other topics of interest. Over the course of the exploratory trip and fieldwork the affiliation was beneficial in sharing across academic scholars on practical tips and insights for conducting efficient and effective fieldwork. The working relationship that was established with the Director of ICAIOS at this time proved vital in understanding the local research context, potential avenues for data sources, relevant contacts, and in ensuring the cultural sensitivity in research design. Contact was also established with the Tsunami Disaster Mitigation & Research Centre (TDMRC), also in Banda Aceh. Primarily focused on long-term urban risk reduction, the TDMRC is working with the challenges of relocating communities who are refusing to move. Consequently, several of the databases housed at the TDMRC were valuable in contextualizing the research.  One week was spent at The Gadjah Mada University (UGM) in Yogyakarta, allowing for broader insights on disaster risk reduction and recovery in the Indonesian context.  Through a guest lecture to a group of Master?s students in the Urban and Regional planning program, feedback on the research project was obtained. A relationship with the department was maintained throughout the course of the fieldwork.  While in Aceh for the exploratory trip, accommodation was provided by a host family in a neighborhood that had been affected by the tsunami.  This allowed for quick immersion into Acehnese culture and way of life. The experience enabled preliminary insights on human dimensions of the recovery process and personal stories of the post-disaster period. It became clear that the general local perspective on recovery was that NGO projects were unsustainable and lacked long-term planning or outlook.  The experience also allowed for the practice of beginner level Bahasa Indonesia language. (Classes in Bahasa Indonesia had been completed at the University of British Columbia over the academic year 2009 to 2010.)   3.3. Field preparation Based on insights from the exploratory trip, research questions were refined and next steps in research design were finalized back in Vancouver, Canada. Upon successful defense and committee approval of the research prospectus, preliminary instruments for data collection and English consent forms were developed.  Compliance with the University of British Columbia?s ethical standards and  57  requirements was ensured. To do this, the research followed the TCPS (Tri-Council Policy Statement for the Ethical Conduct for Research Involving Humans) application process and was reviewed and approved by the UBC Behavioural Research Ethics Board (BREB).  Ethics approval was obtained before departure for Indonesia. In order to enhance working knowledge of the local language prior to embarking on fieldwork, an intensive Bahasa Indonesia language course was completed over two weeks at Realia Indonesian Language and Culture Centre in Yogyakarta prior to arriving in Banda Aceh.  3.4. Fieldwork Fieldwork was conducted over a period of five consecutive months (March 2011 to July 2011). Soon after arrival in Banda Aceh, a posting for a research assistant (RA) was circulated at the Syiah Kuala University campus. Interviews for the position were conducted, an appropriate RA was selected, and consent letters and forms were translated into Bahasa Indonesia at this time.  3.4.1. Targeted case selection Case selection followed a purposive sampling approach, aiming to cover a representative sample of resettlement cases at the same time.  Cases were intentionally sought because they met criteria for inclusion in this study (Palys & Atchison, 2008). The initial criteria developed for selecting cases was based on combinations of dimensions of place and of planning and on the basis of similar pre-disaster conditions (i.e., reported level of development prior to the disaster was similar across villages). This included ?moved? (i.e. villagers were moved to a new location) and ?not moved? (i.e. villagers stayed in their original village) and ?planned? and ?not planned?.  The criteria was later modified to ?moved? and ?not moved? and ?participatory? and non-participatory? and will be explained below.   The reasons for selecting initial dimensions of place and planning were the standpoint that: 1) physical location was a core element of place; and that 2) during the exploratory trip it was observed that planned settlements appeared to embody more place attributes than unplanned settlements. The criteria would effectively cover a representative sample of resettlement typologies while ensuring adequate variation across the placeness variable. Therefore, cases would include an equal number of those of resettlement in pre- disaster locations and resettlement in other than pre-disaster locations in relation to whether resettlement planning was planned or unplanned. Notions of a planned settlement  58  included the enforcement of some level of controlled and organized spatial planning.  The reason for incorporating pre-disaster conditions into the selection criteria was to allow for comparisons across cases in both collection procedures and future analysis techniques.   Five communities that had been tentatively selected during the exploratory trip were re-visited at the start of the fieldwork.  Through consultation with local scholars and field practitioners, a number of additional possible case studies were explored and also visited.  In total, twenty villages were observed in the sampling frame and around fifteen villages were narrowed down as options for detailed study.  Based on direct observations, ten were selected for further surveying. In each of these villages surveyed, initial contact was made with the village chief.  Through informal discussion with village chiefs, and by asking local villagers, a better understanding on the background of the village was gained and the suitability of the case was evaluated in terms of selection criteria, openness to research, and potential for interesting research findings.  For example, one of the Compounds where preliminary interviewing was conducted was a new compound built in Thurbe, located within the fields and mountainous area of Jantho (~40km directly South of Banda Aceh).  The majority of residents were from the island of Pulo Aceh or from the district of Aceh Jaya.  Many of the residents spoke of returning to their old regions, but of prevailing trauma holding them back.  However, this case was not pursued do to difficulties in accessing the community on an on-going basis (i.e., its remote location).  In most cases, the village chief was located through inquiry among local villagers. In some cases however, a meeting with the village chief was organized through a mutual local acquaintance.  Through conversations with village chiefs, it became apparent that the dimension of planning (planned vs. unplanned) was perhaps not the best marker for case selection criteria.  In fact, many officials argued that there was only one true case of a planned resettlement in a pre-disaster location.  This case was classified as the ?ideal? village reconstruction case, having undergone significant land consolidation. It later became apparent that this case was the ?go-to example? to display reconstruction to tourists and had been researched by many scholars.   In addition, it became clear through discussions that although the majority of resettlements not in pre-disaster locations were planned, they embodied very different attributes of place (i.e., some had paved roads connecting houses, other had designated public gathering places, and so on).  The dimension of participation began to emerge as one that seemed to enable more variation in place attributes. Keeping in mind perspectives that a first criterion in selecting cases should be to maximize what can be learned (Stake,  59  1995), the participation dimension seemed one that would potentially make for more insightful case studies, as it allowed for more depth on the resettlement process as opposed to the planned/unplanned dimensions.  Therefore, the selection criterion was refined to combinations of dimensions of location and dimensions of participation.  Other dimensions such as size and proximity to the coast were not considered, as using these dimensions limited the scope of the exploration to more than what was desired.   Villages were classified as ?participatory? or ?non-participatory? based on initial village surveying. During the surveying, village chiefs were asked about the nature of the resettlement process and level of involvement by community members.  In some cases, the village chief invited a community member who played a role in the reconstruction process to join the conversation.  In one of the villages, the chief was previously a housing supervisor hired by the NGO working in the village. Those villages that were described as having any level of involvement by community members during the resettlement were grouped as ?participatory?.  Those where villagers were described as simply being allocated a resettlement site and/or returned to the village upon completion of housing were grouped as ?non-participatory?.  These classifications were supported by reports published by NGOs that had worked in each of the respective villages (i.e. NGO website articles on reconstruction activities).  In Chapter Four, the villages are later ranked on a participation spectrum based on detailed data collected on the resettlement process (see Table 4.5).     Two cases of ?moved?, ?non-participatory? were selected for study as both had the potential for interesting findings yet embodied very different community dynamics. In fact, it was hoped that two cases in each ?cell? would be studied and as such, eight possible communities were selected and surveyed.    However, due to the amount of time needed to diligently study each community and limited resources to effectively do so, ultimately five communities were studied (as listed in Table 3.1.).  ?Resettlement not in previous location? indicates that the community was either relocated to a new location or it was a new community created from scratch in a new location. ?Resettlement in previous location? indicates that the community was rebuilt in the same location as it was prior to the disaster.  The final five cases selected were Neheun Compound, Panteriek, Lampulo, Gampong Baro and Bitai.  Figures 3.3 and 3.4 below situate these villages on the broader map of Banda Aceh and surrounding Aceh Besar.   60   Figure 3.3. Village cases on map of Banda Aceh and Aceh Besar.  Villages indicated as: 1) Panteriek Compound 2) Bitai 3) Lampulo Source: Map provided by BAPPEDA Banda Aceh during fieldwork 3 2 1  61   Figure 3.4. Village cases on map of Masjid Raya subdistrict of Aceh Besar  Villages indicated as: 4) Neheun Compound 5) Gampong Baro 6) Old Gampong Baro  Source: Map provided by BAPPEDA Aceh Besar during fieldwork  Table 3.1. Case criteria and communities selected Three of the case studies (Neheun Compound, Panteriek and Lampulo) were confirmed within the first month of fieldwork. The remaining two cases (Gampong Baro and Bitai) were confirmed during the third month of fieldwork.  Cases were studied from March to July 2011, one after the other, to allow for full immersion into each case and to build confidence and acceptance Participatory Non-participatoryDid not Move(Resettlement in previous location)Lampulo BitaiMoved (Resettlement not in previous location)Gampong BaroNeheun CompoundPanteriek Compound4 5 6  62  with villagers over time.  Approximately one month was spent on each case, with instances of some days of overlap during the span of fieldwork.  This was possible due to the close proximity of cases, with the furthest case being less than an hour away from the author?s place of residence in central Banda Aceh during fieldwork.  3.4.2. Data collection  Data were collected on the three main variables of concern: recovery, resettlement, and placeness. Multiple data sources were used for collection. Recovery data relied primarily on semi-structured interviews at the village level, resettlement data on village surveying, and placeness data on focus group discussions with villagers.  Each data source, however, provided information to some degree on all variables of interest, with some sources more so than others.  This allowed for more comprehensive data collection and has enabled construct validity with data converging in a triangulating manner (Yin, 2003).  Open-ended interviews were also conducted with key experts (i.e., staff at local government agencies, non-governmental organizations and academic institutions) for contextual and other insights. The detailed numbers of interviews and/or discussions for each collection method, along with ranking of data collection by variable are listed in Table 3.2.  Whether the data source was a ?central? source of data or 'subordinate? source of data is also indicated.  Most of the questions and data collected on that particular variable was from the source indicated as ?central?.  Other sources that provided information toward a particular variable are listed as ?subordinate?. The table indicates how central and subordinate sources were used to validate, verify and increase the credibility of information as a technique of triangulation.  Table 3.2. Data source by variable Source: Fieldwork  D a Sou c Number Recovery Resettlement Placeness ContextualS mi- tructu ed interviews 75 Central Subordinate SubordinateDirect bservations N/A Subordinate SubordinateSecondary sources N/A SubordinatePreliminary surveying 10 CentralFocus group discussions 14 Subordinate CentralOpen-ended interviews 26 CentralTotal 125Aspect of Recovery 63  Before each interview and focus group discussion, the purpose of this study and voluntary nature of participation in this study were explained as per the UBC ethics protocol.   A copy of the consent form in Bahasa Indonesia was provided.  Either verbal or written consent was obtained. Verbal consent was a preferred method in cases where community members were hesitant to sign their name to a document, but were agreeable to providing the interview, and in the more informal conversations that occurred over the course of fieldwork.  Interviews and focus group discussions were conducted in Bahasa Indonesia.  The same research assistant was present as a translator for all of the interviews and focus group discussions that were conducted.  At first, interviews were tape-recorded.  However, due to discomfort displayed by interviewees, the decision to take hand-written notes instead was made during the first case study and applied for the rest of the cases. Recovery data Key informant interviewing was chosen to collect recovery data for its effectiveness in gathering broad data, its advantages in being able to access information otherwise unavailable, and its ability to enable an understanding or interpretation of cultural information (Crabtree & Miller, 1999).  The technique was a preferred strategy in the context of village-level fieldwork as it allowed the interviewees to express their views in their own words.  The method has been successfully used to understand perspectives of recovery by several disaster researchers; for example, Berke et al (2008) who, through key informant interviews in six villages of post-tsunami Thailand, found that aid programmes may have prevented the potential of social capital for collective action toward positive outcomes.   Since any one informant would not be able to accurately assess all aspects of community recovery, many informants with different perspectives were interviewed to obtain a holistic understanding of community recovery.  Informants were selected based on their role in the community and ability to comment on a specific aspect of recovery.  Key informants interviewed included, imams19, school principals, health professionals, elected members of the village                                                    19  Imam has been translated in the glossary as a religious leader.  In the context of villages in Aceh, the imam plays a key role in leading communal prayers in the village.  He is also responsible for maintaining the religious gathering space and its congregation (for more information see Graf et al., 2010).   64  government20 and other individuals in leadership roles21 in the community. Key informants were recruited through site visits and networking.  Many of the informants were identified through consultation and introduction from the village chief, while others were selected through introductions facilitated by villagers and snowball sampling techniques. A complete list of informants, along with the number of each type of informant interviewed in each village is displayed in Table 3.3.                                                      20 Members of the village government interviewed included for example, the village chief, compound chief, village secretary, block chief and/or alternative to block chief.  In the cases studied, the compound chief position only applies to Neheun Compound.  Due to the large number of new compounds constructed in the village of Neheun, the village governance was modified to introduce compound chiefs who would function to support the village chief in his duties.  In the case of Neheun Compound, the Neheun village chief carried out only administrative duties; the Neheun compound chief addressed all other concerns.  The village secretary sits under the village chief and plays a strong role in village governance.  For example, in Bitai, the village secretary was active in administrative duties that were not family with the village chief.  In cases when the village chief was killed by the tsunami, the village secretary would step into the role.  Each designated block (i.e. area) in the village had block chief responsible for residents of the area.  In the case where a block chief was not available, an alternative to the block chief was interviewed, based on recommendation of the village chief.  21 Other individuals in leadership roles included for example, the youth leader, women?s group leader, and village elders.  The youth leader in each village is an elected member of the village governance and was responsible for carrying out duties specific to the youth population (i.e. up to around age 35) of the village (for example, social gatherings).  The women?s group leader was in charge of conducting activities specific to women in the village (for example, home industries, socio-cultural events, and so on).  In some cases this group was more active than in others.   A village elder was a senior member of the village (in age).   65   Table 3.3. Key informants by village and category Source: Fieldwork Semi-structured questionnaire templates specific to key informant category (e.g., community leader, health professional, educator) had been developed prior to arrival in Indonesia.  Soon after arrival, templates were tested with local graduate students and scholars housed at the ICAIOS.  The interview protocol was refined to accommodate for cultural sensitivities, context, and to better reflect local vocabulary. Once a suitable first case was confirmed (which happened to be the one that had been selected during the exploratory trip), the questionnaire was piloted and minor modifications were made during the case. Some of these modifications were in making the questions specific to informants, listed in Table 3.3 above. In many instances, the interview guide was adjusted during interviews to accommodate for discussion that emerged.   Informant Totals Village Chief 1 1 1 1 1 5Compound Chief 1 1Village Secretary 1 1 1 1 1 5Block Chief 7 4 2 3 4 20Alternate to Block Chief 1 1Imam/Deputy Imam 1 1 1 1*4Youth Leader 1 1 1 1 4Doctor 1 1Midwife 1f1f1f1f2 6School Principal 1 1f1f3 6Teacher 2f4f1f2 9Women's Group 1f1 2Ex-Chief 1 1 2Village Elder 1f2f +1~4Academic 1 1 1 3NGO Staff 1 1 2Totals 17 17 9 16 16 75         f) Female participant         +) Also serves as Imam/Deputy Imam         *) Also serves as Village Secretary         ~) Also serves as Ex-ChiefBitaiNeheun CompoundGampong BaroLampuloPanteriek Compound 66  Interview questions usually started by asking the informant about their position, their role in the community, how long they had been a part of the community, and their participation in the resettlement and recovery process (see Appendix One). Questions on dimensions of recovery specific to the category of the informant were then asked.  (For example, an educator would be asked about changes and data on school enrolment levels over time, and a midwife on health services and number of reported illnesses over time). An attempt was made to ask questions in a systematic way, in order to allow for comparison across cases.  For example, words such as ?more? and ?less? or ?better? and ?worse? were used when asking about changes.   The length of time of interview ranged from 30 minutes to 2 hours.  Most of the interviews were conducted in or outside the interviewees? place of residence.  In the case of professionals, the interviews were conducted at their place of work (i.e., school, health clinic, village office).  Direct observations on a number of recovery indicators (for example, hectares of productive cropland, number of food vendors, and so on) were noted in each case.  Observations served as a valuable source of data for recovery given the sparseness and/or consistency of reliable and/or accessible data. Detailed observations on specific recovery indicators in each case were recorded following a standard reporting style, for example on quality of roads and houses. Possible new indicators for recovery were explored based on reflections of what was seen (for example, number of renovations).  Observations were listed in a way that would allow for comparisons across cases (for example, tables comparing the cases). Observations were recorded during the same timeframe that interviews were being conducted in each village.  In the last few days of fieldwork, all of the cases were re-visited to make further remarks and to modify existing notes by adding comments on comparable observations. Secondary data that was related to understanding community recovery were also gathered during fieldwork. At the village level, documentation included village office records on population, marriages, births, deaths, school enrolment, health visits, and so on. Reports on the current state of development, including the current village plan, were photocopied.  Some documentation indicative of recovery was accessed at the various government bodies and agencies that were visited, for example, village-level statistical data from the Badan Pusat Statistik (Central Bureau of Statistics) and village-level budget documents from the Badan Pemberdayaan Masyarakat (Community Empowerment Board).  The reliability and validity of documentation was explored through consulting with the person in charge of collection at the respective agencies.  67  Chapter Five on ?Assessing Community Recovery? will include details on how specific recovery indicators were developed prior to and modified during the course of the fieldwork.  These indicators (i.e., marriage rates, reported illness, school attendance levels, etc.) were based on a combination of interview data, observations and secondary sources. Chapter Five will include which specific indicator information was observed and/or probed through the different data collection techniques, and will also describe development of the multi-dimensional recovery index. Resettlement data Resettlement data (i.e., whether it was a resettlement that was not in its previous location or a resettlement that was in its previous location) were primarily obtained during village surveying when selecting cases.  The initial purpose of data collection on this variable was to categorize and select cases for research.  Details on the collection have been described in the earlier section on ?targeting village cases?.  Classifications of moved, not moved, participation and non-participation were based on data from the village chiefs and official documentation like reports and maps obtained from government agencies.  Insights from semi-structured interviewing with key informants as described above provided further details on resettlement process.  As will be described in the next section, in the focus group discussions, many additional challenges and details related to the resettlement process emerged. Placeness data Placeness data (such as sense of belonging) were primarily obtained through focus group discussions with village members.  Focus group discussions were conducted in each of the cases as indicated in Table 3.2 earlier.  The method was used for its effectiveness in understanding a community?s beliefs, ideas, attitudes and opinions of their community (Dawson et al., 1993).  Focus group discussions provided the opportunity to witness rather than influence the community?s interaction on topics within a relatively limited timeframe (Morgan, 1988).   The primary purpose of the discussions was to gain a deeper understanding of place, community life and dynamics in the village, to gather insights on the role of place attributes during the recovery process and to understand current conditions and concerns of the community. Two standard focus group discussions were conducted in each of the cases ? one with males, and one with females.  The genders were separated in order to accommodate for cultural sensitivities and to enable  68  a greater level of comfort among female participants. Additional focus group discussions were held in some of the villages based on the uniqueness of the case, as described in the next chapter.  A complete list of focus group discussions conducted in each village in displayed in Table 3.4.  A discussion template was kept at hand (see Appendix One), though the discussion topics were kept somewhat flexible depending on what the group wanted to talk about. In the spirit of acting as a facilitator rather than interrogator, topics were probed only as and when appropriate.    Table 3.4. Focus group discussions by case Source: Fieldwork Focus group members were recruited through different ways in each village. Though a standard approach to recruiting participants would have been preferred, the dynamic realities of the field made this difficult.  Recruitment included sessions set up by village or block chiefs and/or their wives; sessions that followed attendance at community events; and sessions organized after village gatherings, for example, following weekly Quran recitation classes.  Focus group discussions ranged from one to three hours. Attendance averaged around fifteen to twenty participants, though discussions ranged from five to twenty-five participants.  The discussions were held in public spaces, such as the village mosque or other prominent gathering place in the village. Efforts were made to try to attract an assortment of ages in each focus group, from young adults to elderly.  When composing groups, attempts were made to minimize differences in power and status in order to properly observe Village Group Males/Females Description1 M Newcomers2 F Newcomers3 M Originals and Newcomers4 F Originals and Newcomers5 M Originals (Youth)6 M Newcomers (Youth)7 M Originals8 F Originals9 M Originals and Newcomers10 F Originals and Newcomers11 M Newcomers12 F Newcomers13 M Originals14 F OriginalsOld Panteriek CompoundBitaiNeheun CompoundG mpong BaroLampuloPanteriek Compound 69  dynamics. For example, in one of the cases, discussions with newcomers were carried out separate to discussions with original community members.  When discussion was being dominated by a handful of people, efforts were made to encourage discussion and indirectly channel discussion to other members in the group.  During the discussion, the researcher and research assistant also recorded notes on group dynamics. In many cases, the discussion would be followed by smaller informal conversations. These smaller conversations provided insights on the emotional feelings community members had for their particular ?place?. Similar to the recovery variable, direct observations on placeness indicators was taken in each case. Detailed observations on place attributes in each case were recorded following a standard reporting style, for example on public space and meeting places. Observations were listed in a way that would allow for comparisons across cases. Observations were recorded during the timeframe that interviews were being conducted in each village.  In the last few days of fieldwork, all of the cases were re-visited to make further observations and to modify existing notes will adding to notes on comparable observations.   Data from the semi-structured interviews with key informants described early also provided some detailed insights into placeness indicators. Contextual data In addition to data collection for the three variables described, open-ended interviews were conducted with key experts in order to better understand the state of recovery in Aceh as a whole.  Experts were from government agencies, non-governmental organizations and academic institutions. Some interviewees were targeted to provide specific insights on queries that emerged during fieldwork (for example, UNDP Program staff, to understand disaster risk reduction programming in villages).  This was especially valuable, as the NGO voice was not obtained or incorporated in understanding recovery in each of the cases (for reasons that include difficulty in contacting ex-pat NGO staff that were no longer situated in Aceh).  While the NGOs that were interviewed for contextual data did not directly work in any of the village cases, insights from those interviews were valuable in understanding the similar conditions and structures in which they were operating.  Other were consulted to gain access to pertinent data (for example, BAPPEDA staff, to gain access to GIS maps), to achieve clarity on timeline of resettlement processes (for example, BRR staff) or to gain an understanding of the broader context in which recovery was taking place (for example, the Head of BAPPEDA).    70  Informants were recruited through networking, mutual acquaintances or by visiting their respective offices.  A complete list of informants is displayed in Table 3.5 below.  Interview questions usually started by asking the informant about their position, their perspectives on the recovery process, and insights into priorities during early recovery, mid to long-term recovery and those at the present time.  This was often followed by specific queries or kept open-ended to discussion that followed.  Some impromptu interviews were conducted when presented with the opportunity, resulting in stimulating discussion and unique insights. Open-ended interviews ranged in time from 30 minutes to 2 hours.  Most of the interviews were conducted at the agency office.  Type Position Organization/Agency Data Collected1Vice-Mayor fMunicipal office, Banda AcehCurrent priorities and vision 2Research and Development OfficerBAPPEDA  Provinsi Recovery process3Secretary, Aceh Development ProgramBAPPEDA Provinsi Recovery process4 Technical Support Officer BAPPEDA Provinsi Recovery and housing process5 Unit Officer, GIS Unit BAPPEDA Provinsi Access to maps, guidance on cases 6 HeadBAPPEDA Banda AcehBarracks Coordinator (Past)Recovery challenges, future plans 7 Unit Officer, GIS Unit BAPPEDA Banda AcehAccess to maps, guidance on cases8 Secretary BAPPEDA Banda Aceh Housing verification process9Program Staff fBAPPEDA Banda Aceh Recovery process10 Head of Development BPM Banda Aceh Access to data, insights on budgets11 Budgets Officer BPM Banda Aceh Budget insights on village cases12 Head of Section PU Banda Aceh Current infrastructure challenges13Chief, Reconstruction DepartmentPU Banda Aceh (Current) BRR-Infrastructure (Past)Recovery processG v r ment / NGO14Communications SpokesmanBRR (Past)Local NGO (Current)Lessons learned from recovery15Manager, Housing and Beneficiaries DivisionBRR (Past)Lecturer (Current)Recovery and housing process16Social, Cultural and Religious AffairsBRR (Past)Lecturer (Current)Recovery challenges         f) Female participantGovernmentGovernment / Academic 71   Table 3.5. Experts interviewed Source: Fieldwork 3.4.3. Journal As indicated, a journal was maintained throughout the course of the field work.  Daily reflections were reviewed on an on-going basis in order to help shape the research process.  At times, experiences from one day would inform how data collection would take place the next day.  An on-going analysis was conducted during the period of fieldwork to assist in understanding what to probe or inquire on when there was confusion or uncertainty.  3.5. Fieldwork challenges and ethical issues  3.5.1. Researcher ? research assistant relationship Having a local research assistant present during the fieldwork was vital in attaining the data that has been collected.  To observe cultural sensitivities associated with close and extended interaction across genders and on the basis of personal comfort, the research assistant selected was female. Previously, she had worked as a local hire with a leading donor agency during the tsunami response and early recovery period.  This was highly valuable during the data collection as her background knowledge of the recovery process was available at hand.  Through her work experiences she was familiar with conducting focus group discussions.  At the time of fieldwork the research assistant was considering further studies and awaiting scholarships for a graduate program.  The arrangement that was nurtured between research assistant and researcher was a mutually Type Position Organization/Agency Data Collected17 Staff Member fUNDP Recovery challenges18 DRR Team UNDP Risk reduction programming19 Staff Member World Bank Recovery process20 Head European Commission Recovery process21 Staff Member European Commission Current priorities and challenges22 Head IOM Resettlement challenges23 Head Islamic Relief Recovery and housing process24 Head of Research TDMRC Current research priorities25 Lecturer IAIN University General insights on recovery26 Lecturer Syiah Kuala University General insights on recovery         f) Female participantNGOAcademic 72  beneficial one, for she was keen on improving her English and learning about conducting research through the assistantship experience, and her skills in translating and assistance were vital for carrying out this research.   The vast majority of interviews and focus group discussions were conducted in Bahasa Indonesia by the researcher with the support of the RA.  Though an understanding of basic Indonesian language had been acquired through UBC and Realia, translation by the research assistant was critical at times.  Challenges that are commonly associated with conducting research in a foreign language were inevitable.  Overcoming challenges included for example, ensuring that:  meanings were not being lost in translation; that the translator was not unconsciously attaching bias to answers; and that heated discussions were not interrupted at the wrong times. To minimize this and to maximize capturing all of the information that had been conveyed, time was spent after each interview and focus group discussion with the research assistant and independently to complete notes and fill in any blanks.  Basic training on co-facilitating focus group discussions was provided and over time, discussions were co-facilitated.   Having a research assistant was beneficial in ways beyond translation.  For example, the assistant was able to grasp crucial contextual information about the local cultural nuances and subtleties of language that may have been missed otherwise. At times, the assistant would screen any questions that may be inappropriate for the situation. For instance, during some interviews we would realize that the informant was an ex-combatant from the GAM movement, with prevailing views on the rebellion described earlier.  As part of the government strategy after the peace accord, several ex-combatants had been integrated into village life through resettlement. In some cases, discussing certain conditions prior to the tsunami would be avoided, including topics surrounding identity.  3.5.2. Local expectations  Because most foreigners who had been in Aceh had been there representing an aid agency, the expectation of assistance from me was apparent in many instances.  In some instances during village surveying, villagers would gather around asking each other ?ada uang?? (?does she have money??).  Within the first minute of meeting anyone, the question, ?where are you from?? was always asked. Being Canadian brought a perception of having money, especially within villages that had received assistance from Canadian NGOs and agencies. The need to immediately and clearly communicate intentions for being there became of utmost importance. Once the message that the  73  purpose of presence was for research being conducted by a student was understood, word travelled quickly and expectations of aid were removed. Travelling in a discrete vehicle minimized expectations, as most aid agency vehicles were branded.  Having formal written proof of affiliation with the state university was useful in changing the perception from NGO staff member, to student, particularly among key experts.  Some of the villages that were surveyed in Aceh had received significant attention from researchers in the past.  When selecting cases, attempts were made to choose cases that had not yet been researched, in order to avoid repetition of questioning and of data collected.  Some villages surveyed indicated frustration from previous experiences.  An example is one of the compounds in the village of Mataie that had been narrowed down as a good example of a participatory and moved resettlement. In this particular case a group of women spoke up on feeling exploited by researchers, talking extensively about a medical researcher that had arrived in their village to study their health status and had then left suddenly.  ?Sometimes we are too nice.  We are even feeding them.  But we get nothing in return.  We know they got something from us, but we didn?t get anything? (Female, 40s, unemployed).  The same compound had recently been recorded for a foreign documentary, though they were not sure what for.22  The discussion is an example of the ethical dilemmas that surfaced over the course of the fieldwork.  The villagers that expressed frustration were in villages that were not selected as case studies due to negative perceptions on the research from the very start.  3.5.3. Researcher ? participant relationship The relationship between researcher and participant is sensitive in any research project.  As someone who was visibly foreign to the setting for this research study, this was even more so and drew immediate attention from community members.  Due to the influx of international aid workers after the tsunami, most villagers were comfortable with foreigners.  However, at the time of field research most of the expatriate community had left, and there were very few foreigners remaining in the Banda Aceh and surrounding area.  In most cases, the research assistant was key in facilitating the introduction to the community and explaining the purpose of our presence.  The value of education is held in high regard in Acehnese culture, and affiliation with the state university was significant in gaining initial support from the village chief in respective villages. Once the chief                                                    22 The researcher was not able to identify or locate the name of this documentary.  74  showed support, acceptance by villagers followed.  Most of the meetings were held outside informants? homes and/or in public places; over time most knew about the research and drew more and more engaged.  It was common to spend time having tea with informants and family members after interviews.  Similarly, after some of the focus group discussions smaller groups would remain to informally chat on topics outside the scope of the research.  Building trust and rapport in village cases and among participants was a guiding standard in this research (Patton, 1990). Aceh is a predominately Muslim environment, with over 98% of the population practicing Islam and being governed through Islamic law. As the author is a Muslim, being of common faith was a key advantage in conducting this research. Villagers often conveyed trust, based on common beliefs and principles.  The commonality enabled quick acceptance, as community members were able to associate on the level of greatest importance to them.  For example, on finding out about the common faith, informants would release a sigh, followed by ?You?re Muslim! Why didn?t you say so from the start! Then of course you know how it is for us.? (Male, 50s, village leader)  At times it was as though the entire tone of the conversation immediately shifted to being more relaxed. Furthermore, participating in religious services with members of the community at their local mosques brought a distinct sense of unity.  Through sharing experiences with other scholars conducting research in Aceh it appeared that the unique positioning of being a foreigner, yet of the same faith, led to community perceptions and interactions that were quite different than those experienced by foreign colleagues.  At the time of fieldwork in 2011, just over six years had passed since the disaster.  Many recalled the event and aftermath with vivid descriptions. During several of the interviews informants would become teary eyed and emotional.  At moments, continuing the interview would become difficult, and even changing the topic felt insensitive. It was common for villagers to express memories of loved ones who passed away from the tsunami through comments such as, ?my daughter would have been your age now? (Female, 50s, teacher), ?my son didn?t want to be a fisherman? he was planning to study in university like you? (Male, 40s, fisherman). During an interview with one of the village chiefs, tearfully said, ?you look like my wife? you remind me of her, so how can I not help you?? (Male, 50s, village leader).  How to properly handle emotional encounters during the data collection was difficult.  At times it felt wrong to be contributing to the resurfacing of memories.  Nevertheless, being empathic to their losses was important.   75  3.5.4. Benefit to participants It is hoped that the benefits of this research to participants came partly as they realized the importance of their role within their communities.  As many pondered on current community issues and conditions during interviews and focus group discussions, it is hoped that discussion brought to light some of the concerns that could be addressed and prioritized by the community over time.  The majority of survivors in some of the villages were young adults, and in several instances they would comment on feeling inspired by seeing another young adult that they were able to relate with is pursing graduate studies.  It is intended that findings of this research will be shared in Aceh in due course.  One avenue in which this will be pursued will be within the academic community in Aceh, through participation at future ICAIOS conferences and presentations to students and scholars working at ICAIOS. Another avenue will be through some of the agencies and organizations in which relationships have been developed. Several key experts have requested findings to be sent to them so that they may possibly gain insights for future plans.  Some experts requested meetings on preliminary findings and possible recommendations for development prior to my return to Canada.  For example, an informal discussion was conducted with the Head of the Regional Body for Planning and Development, Banda Aceh in one of the last few weeks of fieldwork. 3.5.5. Trustworthiness  It was important to strive toward an academically rigorous study, which includes establishing the trustworthiness of this research through various strategies and criteria (Lincoln and Guba, 1985). Specific strategies were used throughout the research process in order to increase to worth of this study (Krefting, 1990).  The research process (i.e. several months in Vancouver between the exploratory trip and data collection displayed in Figure 3.1) enabled some time for personal reflection on the role of the researcher and positionality in the data collection process.  An extended period of time was spent in each of the villages in order to build a rapport and ensure that responses were built on personal experience of the community members rather than what was expected from community members.  The field journal kept during the fieldwork process (see Section 3.5.4) enabled a process of reflexivity ? an assessment of the researcher?s background, perceptions and interests throughout the fieldwork (Ruby, 1980). The triangulation of both data methods and data sources (see Section 3.4.2) was a technique used to enhance the quality and credibility of the research (Krefting,  76  1990).  Continuous exchange and dialogue with researchers and professors at the ICAIOS was carried out over the course of fieldwork to allow for peer examination and to ensure dependability (ibid.). Finally, neutrality was enhanced and personal biases were minimized through constant feedback and dialogue with the local research assistant who was familiar with qualitative research methods (see Section 3.6.1).           77  CHAPTER FOUR: CONTEXT AND CASE STUDIES Chapter Four presents data collected during the fieldwork process that has just been described in Chapter Three.  Contextual data that was collected during key expert interviews (i.e., From government agencies, academic institutions and NGOs) were analyzed thematically by organizing interview notes within the major themes that were apparent through a preliminary analysis of interview notes.     The first part of the chapter develops an understanding of the current state of recovery in Aceh, including contextual insights relevant to Banda Aceh and Aceh Besar.  It concludes by summarizing some of the challenges and priorities that were expressed as being of major prominence in the immediate and near future.  The second part of the chapter provides descriptions of each of the case study villages explored in the research. Each description has common sub-headings including overview, resettlement process and current conditions.  Descriptors and variables of concern are summarized at the end of these descriptions. 4.1. Transition from recovery to development 4.1.1. Background In order to situate this study, it is important to understand the transitional environment of tsunami-affected Aceh as it moves from a state of recovery to that of on-going development. While much has been written on the response and early recovery in Aceh, research on long-term recovery and development is sparse, particularly after 2009. This is partly due to the presence of a strong international academic community in the years immediately after the tsunami, compared to mere handfuls that were still actively conducting research at the time of this study.  Consequently, part of this study aimed at gathering basic perceptions on this unique period of transition.  Insights on the transitional environment were gathered through interviews conducted with key experts in Banda Aceh and Aceh Besar over the course of fieldwork in 2011 (See Table 3.5.).  With some overlap, the bulk of key experts were local leaders from governmental agencies (n=16) in Banda Aceh and Aceh Besar.  Several others were from local or international NGOs and/or donor agencies that were still operating in the region (n=9), and a small number were from local academic institutions (n=5).    78  Early in their interviews, each of the 26 experts was broadly asked, ?Is recovery complete??  A definition of recovery outlined earlier in Chapter One was provided (the restoration and improvement of living conditions of disaster affected communities). However, in their responses most described recovery simply as a return to a normal and/or acceptable level of functioning.  As listed in Table 4.1 below, few experts responded with a clear response of yes or no (n=8).   Three times as many responded yes (n=6) over no (n=2).   One of the ?no? respondents was vehement in his response, while the other claimed that, ?No, recovery is maybe 95% complete; I think two to four years more are needed for the rest? (Male, 60s, Government)23.   As expected, around three-quarters of the respondents did not give a yes or no, but rather, provided a more complex answer.  What the breakdown shows is that even though over 6 years had passed since the disaster event, recovery in Aceh was more complicated than a simple binary question.   Table 4.1. Responses to question on whether recovery was complete Source: Fieldwork Respondents described the situation at length, some directing the conversation in specific directions, others venting frustrations on the recovery process, while a handful questioned what was even meant by the notion of recovery.  Many alluded to different dimensions of recovery being completed (for example, housing) while other dimensions were neglected (for example, public infrastructure).  Many others divided recovery into short term and long term and referred to the current stage as a nearer completion of long-term recovery.  Even though the question inquired on ?community? recovery, the response would either refer to individual recovery, or the recovery of Banda Aceh or Aceh Besar in their responses.                                                      23 Key expert interviewees will be referred to in this way (i.e., Male/Female, Age, Type). Key informant interviewees will be referred to as (Male/Female, Age, Village), except in the ?Descriptions of Communities? in Section 4.2 where they will be referred to as (Male/Female, Age).   Y s No Other (Examples below) ?Well, depends?? ?It?s not such an easy question...?[Laughter]? ?Good question??What is recovery??6 2 18 79  To get a general sense of the before and after of the context from the perception of key experts, each of the 26 experts were asked, ?In your opinion, is the current situation better, worse, or the same as before the disaster?? Responses are indicated in Table 4.2. below.  While an interview template was not used for key expert interviews, this initial question was used primarily to trigger discussion, which in most cases pointed to key challenges faced by the region. These key challenges will be described in the discussion that follows and summarized in Section 4.1.5.    Table 4.2. Responses to question on state of current situation   Source: Fieldwork The question yielded some interesting answers amongst the respondents. For example, ?Banda Aceh is now much better.  Actually, it is 10 years better than it is supposed to be.  It looks like a new city now? (Male, 50s, Government).   It became clear during interviews that a multitude of unique historical, political, economic and social issues had influenced the process of recovery in Aceh, and the path of development that Aceh was transitioning into.  To go into substantial depth and discussion of these and other key elements of transition is not the purpose of this section or study.  Rather, select excerpts from an analysis of key expert interview findings are presented to provide contextual overviews and insights into major observations that are important for this study and will feed into further data analysis.  Findings are grouped under three dominant themes that emerged as being of key importance to development during the transitionary phase of recovery. These are: the aftermath of international aid, alternative approaches to urban economic growth, and societal changes.  4.1.2. The aftermath of international aid Closure of the Agency for the Rehabilitation and Reconstruction (BRR) April 2009, four years after the disaster, marked the closure of BRR and the transfer of responsibility for overall coordination and sustainability to the Indonesian National Development Better Same Worse19 0 7 80  Planning Agency ? BAPPENAS (ADB, 2009).24  Although long anticipated, all of the international and local NGO key experts interviewed alluded to negative implications of the BRR closure, claiming it was far too soon.  As mandated by the closure, the vast majority of the international NGOs were required to quickly wrap up projects and leave the region.  As one NGO informant described it, ?in 2009 everyone geared down and ran away? (Male, 50s, NGO). In many cases, NGOs entirely abandoned incomplete projects while others left projects to untrained local staff. Challenges in visa re-entry and extended stay led to narrowing the scope of many projects and premature project monitoring and evaluation. Excerpts from interviews are included below:   With the closure of BRR, many problems with work permits and visa extensions surfaced.  Strict processes were imposed by the Indonesian government in early 2009... and these were implemented by the military intelligence.  All of this led to early wrap up for many projects.  They [BRR] chose to cut everything? it didn?t matter if the project was finished or not.  It is very sad that BRR closed so prematurely.  A lot of NGOs would have continued had BRR stayed open.  Even worse is that many thousands of locals suddenly became unemployed prematurely with the closure. (Male, 60s, NGO)  Because of the pressure put by BRR, NGOs were quick to finish their work.  BRR said to us you have to finish by 2008 or we will take over the project.  Otherwise NGOs would have stayed for 5 years or more ? that was the initial plan for most of us. (Male, 50s, NGO) After the BRR closure, things have become very difficult for donors.  In 2010 there were new regulations enforced, so now donors must coordinate with the Government of Indonesia and the relevant Aceh government body (Dinas) on how the funds will be dispersed. (Male, 50s, Government) According to official documents, prior to their closure, BRR had developed an exit strategy in which assets would be transferred from BRR and other stakeholders to various government bodies in order to ensure continuity between the reconstruction phase and longer-term development (ADB, 2009). However, based on interviews, there were substantial gaps in this process. As one NGO representative put it, ?Before closing, BRR gave some homework to the government.  But the government cannot finish the homework? (Male, 30s, NGO).  According to a government informant, many transfers were still pending ?We still have to finish the handover of projects like public                                                    24 A transitionary agency, the BKRA was established by the Government of Indonesia for the period of 2009 as a body in charge of the coordination, synchronization, and the monitoring and evaluation of reconstruction activities, and to prepare ongoing development plans.  At the time of fieldwork, BKRA had shut down, and few informants raised the BKRA in discussions as having played a significant role beyond the coordination of remaining construction activities.  81  facilities and government facilities. Only the housing part is done? (Male, 50s, Government). Several key challenges in the current day were attributed to the inefficiencies and deficiencies of various transfers.     A big issue now is asset transfer and maintenance.  Before BRR closed, it should have been clear to designate the particular ministry or agency that was responsible for each project? even in the cases they did do that, they needed to have made sure that sufficient funds were set aside by that particular designate.  Things are gradually deteriorating now? only a small percentage of infrastructure projects are being maintained. It will only get worse over time. (Male, 60s, NGO) Numerous topics that emerged in interviews related to complications during the reconstruction process and unique challenges that had to be overcome. These complications included: unclear land borders, inconsistent quality in building materials, corrupt contractors, beneficiaries getting multiple houses, unwarranted newcomers demanding houses, and unreliable data.  For example, a key player in the reconstruction strategy commented, ?Housing, land, statistics, all of that data was very dynamic during the recovery process? (Male, 60s, Government).   Many interviewees spoke at length of housing-related difficulties and possible lessons learned for future events.  For instance, one government official spoke of too much being built too fast: Immediately after the tsunami, it was hard to coordinate the NGOs, what they were doing and where, who was to get a house and why.  It should have been controlled better and the rebuilding along the coast should not have happened? (Male, 50s, Government).  This practice was described as being common because it was easier for NGOs to rebuild on the same land (due to security, legality, and so on), which discouraged them from moving communities to safer ground.  Several researchers (i.e., Clarke et al. 2010; Lizarralde et al. 2010; Fitzpatrick 2007; Verby 2007) who conducted research earlier on in the recovery process have studied many of these complications, including housing, in significant depth, and for that reason within this study related data will not be presented in much detail.  What is important to note, as repeatedly asserted by informants, is that with the housing sector now considered a ?closed matter? and a ?private issue? there was little work taking place in terms of new housing for survivors in 2011, even though many related matters remain unresolved. For example:  82  [On recovery]  Housing is complete ? in fact there are too many homes now.  Education is complete ? we have some of the best quality schools in Indonesia.  Infrastructure though... like roads and ports? they still need support. (Male, 30s, Government)  Before, almost every day there was a demonstration here about housing.  In 2008, we made a housing verification team for the city of Banda Aceh. But until now the issues of double houses is not solved.  Some beneficiaries didn?t get a house.  Some ended up getting two or three.  We have identified many cases like this and have given the cases to the police, but the file is stuck there and there has been no action done on this yet. (Male, 40s, Government) Rather, current (2011) infrastructure priorities in Banda Aceh have shifted from housing and large-scale projects to small-scale infrastructure projects.  These include road widening plans, and work on community water and electricity provision.  As one informant, who explained the BRR process at length, said, ?In most villages, first it was the village planning, then the housing, and then they think about water and sanitation.  They should have thought about housing and sanitation at the same time? (Male, 50s, Government).  Many government officials pointed to challenges in basic infrastructure provision across many communities and alluded to a recent focus on developing public facilities such as water systems, drainage supplies and sanitation systems on a community scale.  Funding mechanisms from BAPPEDA had also shifted toward supporting these priorities, with one funding mechanism requiring villages to use 70% of funds toward infrastructure development. Remaining INGO Presence After the BRR closed in 2009, international NGOs were forced to shift their programming from recovery to development through the launch of the Aceh Economic Development Financing Facility (EDFF), through funding from the Multi-Donor Fund (MDF).  Set up by 15 different donors (including the European Union, Canada, World Bank, Asian Development Bank, and so on) to support the government of Indonesia during the recovery and reconstruction, the MDF was one of few visible active international bodies remaining in Aceh at the time of interviewing in 2011 (with exit plans for June 2012).  Insights obtained from the MDF reinforced current priorities and activities being expressed by local NGOs interviewed.  These priorities centred around directing programs to regions outside of tsunami affected regions.  For example, the focus of EDFF programs was for development within inland Aceh, with no money being allocated from the program to Banda Aceh or Aceh Besar (i.e., any of the village cases being used in this study).  This was justified by the fact that international aid delivery had widened the gap between coastal areas and interior, with ?the real poor now only being visible in the interior of Aceh - they are really far behind from the rest of Banda  83  Aceh and Aceh Besar? (Male, 50s, NGO).  In addition, many key experts were weary of further development plans.  For example:  Development is going to get worse, and is already getting worse, even compared to before the tsunami.  It is because of the corruption, especially at the government level.  (Male, 50s, Academic) We are just about coming out of the recovery phase and moving on to the economic development phase - but that needs lots of international support, which has left.  Instead, there is a lot of hope for private investors? (Male, 60s, NGO). Limited capacity building and challenges in livelihood support  On paper and under the premise of effective development programming, capacity building was emphasized as an important component of the recovery efforts in post-tsunami Aceh.  Many livelihood programs were successful; however, the majority of reconstruction efforts were focused on physical reconstruction.  As one interviewee explained: Is recovery complete? Yes, from the physical restoration side it is finished.  But from the true recovery side, economic recovery and livelihood building, it is still not complete.  There is still a lot to do. (Male, 30s, NGO).  While many parts of Banda Aceh and Aceh Besar have new buildings and facilities, deficiencies in capacity remained across local government bodies, sectors and institutions. It appeared that the capacity building that was currently being pursued by the government was largely being done at a provincial level, and not so much at a district level.  Most of the key experts referred to gaps in capacity building:  In terms of infrastructure, we did well.  Human capital is the problem.  The government was too focused on infrastructure? all of the programs put a priority on that.  They are not really focused on soft development?. For example, all the money was spent on school buildings, but not on upgrading the teaching skills of teachers. It really is a shame. (Male, 50s, Academic) All of the infrastructure is done, now we have to focus on how to manage it.  How can the government provide public services in the infrastructure.  We have a very good hospital, clinics, and school - there are more than the national standard now in terms of numbers and quality of facilities.  We don?t have the doctors, teachers and so on? they are not integrated into the system and there needs to be capacity building for official staff. (Male, 40s, Government) A multitude of administrative structures are there in the government, but the awareness of how to use them is not there.  The capacity of the government?s middle management is very  84  poor.  That is the challenge. There were some NGOs and sectors of BRR who were working on capacity building, but it was just not enough.  (Male, 30s, NGO) Implications of the limited capacity extended to current budget limitations. For example:  Don?t get me wrong? government funds received are high, and structures in place are good, but the capacities to handle them are low.  The money has to go back to Jakarta.  You see, if the Aceh government doesn?t spend the money that the national government sends by a given date, the budget money gets taken back. (Male, 60s, NGO)   Similarly, many informants pointed to a limited emphasis on livelihood support during the recovery process and the impact that it was having on the current situation.   The livelihood interventions that were put in place are not working. Yes, sewing machines were distributed and yes, some are being used, but many are just lying around.  How many seamstresses can you have in one village? Same with the distribution of cows ? some people just didn?t know what to do with them. (Male, 30s, NGO) In addition, conversations with research colleagues and interviews with academic scholars pointed to deeper issues in livelihood support that was implemented without paying attention to underlying or prevailing mindsets (i.e., historical and political pasts).  For example:  We tried to do a revolving fund for livelihood development, but it failed.  Why? Because anything that comes from the government, the thinking of people is that this is our right and there is no need to return it.  We need to establish the value first, before giving the money. For example, we give a boat for three families, but they can?t seem to work together.  We need to teach them how to work together first, and then give them the tools.  The situation is not just like this because of the tsunami, though it made it worse, but it goes back to the Soherto25 regime, which was trying to centralize with authoritarianism.  It didn?t give people the chance to actualize their full potential and it destroyed the values of people? Even though people say it is all of the outsiders that have changed the values of people, I don?t think so. It was before with Soherto.  In fact, so many NGOs are doing work on gender, democracy and so on, but it is the government who does not have enough intention to do it?. During the Soherto regime, the leader was served.  We need to change the thinking to leaders serving.  We need a paradigm shift, but for this there is a long way to go.  (Male, 50s, Academic) Many individuals who were contacted ? whether they were key experts, community members or the local butcher ? spoke of increasing difficulty with the job situation across Aceh.  For example, an NGO representative argued that, ?private sector creation is the engine of growth, but in Aceh,                                                    25 Soherto was the second President of Indonesia who held the office for 31 years under his authoritarian regime, before resigning in 1998.  Under his ?New Order? administration, he created a centralized and military-controlled government.   85  after the tsunami, it didn?t happen? (Male, 30s, NGO).  While an informant from the university anticipated negative impacts down the line, ?Job creation is a big problem.  If people continue like this with no job creation, then we will go back to a state of conflict because of no money and no employment options?  (Female, 30s, Academic).  With many locals having been employed by NGOs immediately after the tsunami, now that the majority of NGOs had left the region, unemployment rates were high. While exact unemployment rates and out migration rates are not available, this general observation of few jobs supports the population growth trends outlined in Section Findings also support conclusions in the Aceh HDR, stating that "the need to resuscitate the economy in order to generate productive jobs for all" is among the bigger challenges of the current day (UNDP, 2011, iii).   4.1.3. Alternative approaches to urban economic growth Expansion to the South As described earlier and reinforced by planning agency officials, rebuilding after the disaster commenced before a spatial plan for Banda Aceh and its surrounding area was published. For example, ?According to our first blueprint, only a few people were to be allowed on the coastline.  We needed to have a new spatial plan right away, but there was no time? (Male, 50s, Government). Consequently, many government interviewees referred to the period after the disaster as a ?missed opportunity?.  Planning agencies were trying to make up for the missed opportunity by encouraging movement to the South and away from the coast.      Dialogues amongst government informants in Banda Aceh often related to plans to expand in the South of the city alongside expectations that the economy would grow in that region. Two interviewees even described the South as a future ?sub-city? of Banda Aceh. The land in the South previously belonged to neighbouring Aceh Besar, but an agreement between the mayors in 2008/2009 passed the land over to the city of Banda Aceh. With a concerted effort to make the South more attractive, many government officials were keen to share development plans and detailed PowerPoint presentations at various meetings that were attended during the course of the fieldwork.  These presentations would often precede, accompany or follow key expert interviews. In some cases several government staff would be asked to share detailed plans, which ranged from 5 to 20 year programs.  The Planning Agency?s detailed spatial plan for 2016 included in Figure 4.1 below  86  presents this vision26.  As indicated in Figure 4.1, it was anticipated that in 30 years, the city centre would have fully transitioned to the South.   Figure 4.1. Banda Aceh spatial plan 2016. Big orange circle marks the ?South?. Source: Map provided by BAPPEDA Banda Aceh during fieldwork Reasons for encouraging movement to the South were explained by interviewees under four themes: as a risk reduction measure, as a viable solution to land shortage along the coast, as a means to creating new livelihoods beyond the traditional fishing culture, and in order to achieve ?Aceh Green? goals (described in the next paragraph).  Since the South is a significant distance from the coast (approximately 8 km), the idea is that developments and communities will be at a reduced risk from flooding and/or another possible tsunami event.  However, the Sumatra fault along the south will still pose an earthquake risk. With the loss of land at the coast and the price of land in Banda                                                    26 Note: the spatial plan for 2016 did not have any influence on the communities selected as the case studies for this study.    87  Aceh having increased by tenfold  (according to the City Planning Office), land in the South at the time of study (2011) was amongst the most affordable. However, the key challenge expressed by planners was the lack of incentives for people to move.  Even when relocation incentives are offered (i.e., providing land in the South), it has been largely unsuccessful.  As one key government player in this process asserted: We lost the moment after the tsunami. That is when the relocation could have happened.  People along the coast say it is business as usual.  They seem to have forgotten about the tsunami?  (Male, 50s, Government).  In addition, the expansion also comes at the same time as sustainability initiatives, with a strong commitment to an ?Aceh Green? campaign.  Launched in 2007 by Aceh?s Governor Irwandi as a means to promote environmentally sustainable economic development, ?Aceh Green? is a provincial development paradigm that is actively being mainstreamed into all provincial development planning and consists of many distinct parts including energy security, environmental conservation, waste management, and so on.  One component is manifested as the goal of making Banda Aceh 30% green by 2029, meaning 30% of land in Banda Aceh would become open space owned by the government.  Based on interviews, estimates specific to Banda Aceh point to a current ?Aceh green? level at 8 to 9%.  One government officer explained that the goal would become possible through land acquisition of both personal property and businesses who decided to move away (i.e., turning gas stations into parks).  She argued that ?if the price is okay, the people are agreeing to move? (Female, 30s, Government).  Relocations offered by the government were to the South, including to new apartments made in the South.  The campaign also requires that protected lands not be cultivated, which accounts for 50% of the province of Aceh. Only unprotected areas can be used for economic activity and future growth.   At the time of interviews the South had recently opened a new bus terminal, a new ring road, several new small markets, and new private housing developments. A long-term vision for the area includes the development of a new big market ? with hopes for it to become Banda Aceh?s central market ? and new major roads to link and expand networks in the South. The orange circle at the south of Figure 4.1. is labelled ?Kota Baro?, meaning ?New City?, the arrows describing various evacuation routes (both from the coast and from the fault line in the South), the purple circles indicating key commercial centres (with the newest being within the orange circle), and the yellow circle indicating the new transportation hub (again within the orange circle). Other new infrastructures are indicated through the legend.  However, while the government was pushing for the  88  macro infrastructure planning of Banda Aceh, many informants argued that the micro scale planning was still a challenge and this was not captured in broader spatial planning processes. Tourism In addition to plans to expand in the south was the vision of ?Islamic tourism?. This vision is one that is promoted by the Mayor and municipality of Banda Aceh.  It is the hope that tourism would be a driver to stimulate the service industry to the South.  With the recent establishment of a Culture and Tourism Agency in partnership with government agencies, a Visit Aceh Year 2011 program was designed to promote the tourism industry in Aceh.  It is hoped that tourism will become a significant source of revenue for local administrations.  Many initiatives linked to the tourism agenda are visible across Banda Aceh.  These include large signage, government officers sporting ?Visit Banda Aceh 2011? uniforms once a week, musical performances and cultural parades.   The period of fieldwork was timely as it occurred alongside this initiative, which allowed for participation at these events.  Insights from the events would often explain observations and discussions that would come up during interviewing and field visits.   For example, it appeared that the rule that government officers should wear their official tourism shirts every Thursday was only practiced strictly in one of the village cases (Lampulo), where all of the village school teachers and principals were visibly abiding by this rule. Attending festivals and organized events also enabled informal conversations on priorities of development plans all over Aceh.  This included the ?Banda Aceh City Expo?, an event over the course of five days at the end of May 2011, to enable cooperation across all sectors including between governments and private investors. Distinct initiatives in marketing specific landmarks as Indonesian tourist destinations (for example, Wei Island, commemorative poles, and the tsunami museum), then in their infancy, were taking form.   While many spoke positively regarding the musical and cultural events, the majority of key experts from NGO and academic communities did not speak favourably about the emphasis that had been placed on the tourism agenda in Aceh, believing that other priorities were more pressing.  For example, when discussing the current state of development, an NGO representation argued that: The problem is that our top leaders want to put priority on tourism. But, in reality there needs to be economic priority for local population activities and local livelihoods support. This is a big problem in Banda Aceh. (Male, 30s, NGO)  89  4.1.4. Societal changes Cultural shifts  Tsunami-affected Aceh was described as having experienced substantial cultural shifts since the disaster event.  While some described this as a direct impact of the event and related consequences (i.e., the influx of the international aid community), others attributed the changes to the end of an era of conflict.  Regardless of cause, many of the discussions with key experts across government and academia referred to cultural shifts as being a key element of societal change influencing the course of development in Aceh.   Displacement, newcomers to the region, and new marriages were all referenced as contributing elements of cultural shifts, in terms of new customs, traditions and values.  Furthermore, the implementation of Shariah law (see Section 1.3.5.) soon after the tsunami influenced cultural shifts associated with religious practices, including stricter guidelines and rules on practicing the faith.  Values associated with family had also experienced change.  This was associated with NGO practices that were described as changing family structures and leading to shifts in notions of family.  For example:    Before the tsunami it would be common for there to be three families ? three generations ? living in one house.  But after the tsunami, each of the families acquired their own house, so it has become three independent houses on the same lot.?  (Male, 50s, Government) At the same time, the aid brought changes to attitudes with key experts alluding to increased jealousies, dependencies and thoughts of entitlement among the people of Aceh.  Much of the jealousy was said to have arisen from discrepancies in aid distribution.  For instance:  At first, materials was not so important to people.  Some people were even accepting bamboo homes.  After the second year when materials were more available is when the jealousy started with all of the different houses, with people saying ?Why is their house better??.  In the third year, houses were even better, and in the fourth year, the best houses... God was rewarding them for their patience.  It was not discrimination.  And the renters were the last to get homes, so the renters got better housing. (Male, 30s, Government) Aid schemes, such as 'cash for work' was described as having contributed to expectations that extended to cultural shifts.  Cash for work schemes were essentially programs whereby NGOs would provide cash for community members to clean up debris within their communities.  While positive impacts of these schemes are undeniable (i.e., temporary provision of income), some reported a new culture of not doing anything (i.e., cleaning around their village) for the best interest of their  90  community without compensation for doing this by an NGO. These expectations extended to the job market.  Having been exposed to NGO salaries, expectations from jobs had changed.  This was reported to be especially profound amongst the youth and was supported by village case discussions.  For example: For our ?Cleaning Together Days? before the tsunami there was a much higher number of participants, but now we are getting very few.  Maybe this is a mistake of the NGO system.  Because everything we do, we would get paid.  Now we expect money for everything, and won't do it if there isn't money (Male, 30s, Bitai).  The social community is changing? no one is doing anything for free anymore.  Because we were getting cash for work of 30,000 Rp per day and we have become dependent on that.  Now everyone expects money for helping the community (Male, 30s, Bitai).   Before the tsunami we had ?Cleaning Together Days?.  Then the NGOs brought Cash for Work, but now that is not operating anymore.  People don?t want to go to ?Cleaning Together Days? anymore, they want to get paid for doing it.  We have to get back to doing it for the sake of the community (Male, 40s, Lampulo) Alongside expectations of compensation, dependencies that had grown out of recovery programs were also raised.  There was also evidence of communities finding opportunity that went against recovery initiatives.  For example:  In terms of recovery, infrastructure like roads, highways and ports still need support? In Calang we are having a lot of troubles. You see, the community doesn?t want to move so that we can continue building the road. The community is making money to cross people across the water ? and it is a source of livelihood for them now! Anyone who needs to cross over has to use their raft crossing? it is a good business for them.  (Male, 30s, Government). Institutional transformations From a region where the vast majority of people did not know what a tsunami was, let alone have sufficient preparedness in place, Aceh has become very active in disaster risk reduction thinking and programming.  During interviews with key experts, interviewees demonstrated an institutional shift in mainstreaming disaster risk reduction into everyday thinking and prioritization.  For instance, interviewees from the Planning Agency and Public Works discussed mitigation plans, including one to make a 40km dyke/ring road from the West coast of the city (Uleelee) to the airport.   However, they also mentioned that the land acquisition for making the road was costing more than building the actual infrastructure. Detailed spatial plan indicating evacuation roads were also shared (for example, Figure 4.2 below).  91   Figure 4.2. Banda Aceh spatial plan with evacuation routes. Source: Map provided by BAPPEDA Banda Aceh during fieldwork  As indicated in the legend, relief roads appear in green, escape roads in red, main refuge route in blue and new refuge routes in black dotted line. Key landmarks, including refuge buildings and evacuation towers are indicated by symbols in the legend. Local initiatives supported the institutionalization of a culture of DRR.  For example, interviews with members of TDMRC and staff at the UNDP DRR program provided insights into programs being instituted.  Achievements included initiatives such as annual tsunami drills across the coast (recently attended by 4,000 community members), the incorporation of DRR into school curriculum, and the awareness and use of escape towers.  The escape towers functioned as multi-purpose buildings with some being used as community centres.  The same institutional shift at the  92  community level, however, varied in spite of many similar strong initiatives. Further analysis in the next chapters will explore this and possible reasons in greater detail. 4.1.5. Summary  As described throughout the section, the transitional environment of Aceh at the time of fieldwork was one marked by a number of challenges as it moved into re-establishing a course of development. The most prominent challenges during this transition that have been discussed include the following (in no particular order): 1. Lack of on-going maintenance of new infrastructure built by donors and NGOs (i.e., schools and hospitals). 2. Lack of clear roles and responsibilities of government agencies. 3. Need for the creation of new livelihoods. 4. Limited land availability near the city centre for new developments, and high land prices.  5. Need to incorporate risk reduction into future development plans. 6. Societal tensions due to profound cultural shifts.  As discussed, a number of key priorities were also emphasized by many of the experts and reinforced during involvement and engagement with local activities.  All of these were spoken of as primarily a responsibility of the government, and secondarily as a responsibility of the people. These include (in no particular order):  1. Promoting tourism in Aceh. 2. Working on roads and drainage systems.  3. Instilling a culture of disaster preparedness and prevention 4. Relocating populations to the South of Banda Aceh 5. Capacity building within government agencies These challenges and priorities shape the unique context in which village cases are positioned. While each village case will be described in detail, these broad issues will be important in understanding the broader state of affairs in which they exist.   93  4.2. Description of communities 4.2.1. The compound of Neheun Overview: a non-participatory case of resettlement not in previous location Neheun Compound is a ?moved, non-participatory? case.  Located 15 kilometres outside of the City Centre of Banda Aceh, it is part of the Masjid Raya subdistrict of Aceh Besar (see Figure 3.4).  On what was previously forested land, after the tsunami, the hillside compound was built by NGO-A27 in 2007, without consultation from beneficiaries.  The compound itself goes by many other names including: NGO-A Village, Friendship Village, and Jackie Chan Village (due to of rumors that Jackie Chan had visited the village during the reconstruction).  Though called a village, administratively speaking, it is a compound that is part of the bigger village of Neheun.  To support the Neheun village chief, who is mainly in charge of administrative duties, each compound has its own compound chief allowing for a decentralized village leadership structure. With population estimates of around 12,000 villagers, the original village of Neheun is expansive and dates back generations. There are four other new compounds similar to Neheun Compound; these were all created after the tsunami as a means to accommodate those left without land and/or shelter.  Each compound was built by a different NGO, and each development could be easily distinguished by its unique architecture and uniform roof colours (see Figure 4.3). While some areas mark the original village of Neheun, others, like Promnas Ujungbatee (a government housing site), are pre-tsunami additions off the main village lines. All compound chiefs typically meet with the Neheun chief once a year for the planning and allocating of the budget and under rare circumstances, to resolve any major problems.                                                     27 The actual name of NGOs operating in communities has been kept anonymous (e.g., NGO-A, NGO-B, NGO-C, and so on).  This has been done in order to avoid the misrepresentation of any of the NGOs due to potential sensitive information revealed by community members. Furthermore, all of the information on resettlement was obtained by interviews with community members and NGOs were not consulted on accuracy of the information.  94   Figure 4.3. View of other compounds around Neheun Compound Source: Photograph by author (D. Panjwani) Neheun as a whole includes mountainous, swamp and coastal land.  Neheun Compound itself is located along the mountainous land and is significantly elevated from the coastline (see Figure 4.4.). The compound is frequented by outsiders, as it is a famed resettlement village with breathtaking panoramic views of the ocean. Physically distinct and relatively isolated from the rest of Neheun because of its elevation, the community stands out on its own.   For this reason, although technically a compound, Neheun Compound is used in this research as a stand-alone village case.  95   Figure 4.4. Uphill view of incline in Neheun Compound Source: Photograph by author (D. Panjwani) Documented community data for this case was very sparse; almost all of the data described is primary information collected through interviews.  As described by the village office, due to the dynamic nature of resettlement (i.e., people were still coming and going in 2011) in the compounds of Neheun, the office did not have detailed population data for each compound.  In fact, population estimates varied by the thousands depending on who was asked. The approximate population of Neheun Compound that will be used is 300 families and 1,000 people, as indicated by the compound chief. The same person indicated that around 70% were young adults, with one or two small children; there were fewer than 20 elderly members in the village. The population is mixed Acehnese Muslim and Chinese Buddhist, though a significantly smaller number of the latter. Based on discussion with Block Chiefs, the vast majority of males in the community were fishermen, fish sellers, builders, or becak28 drivers (pedicab). There are very few government officers that live there (~less than 5%).                                                     28 As translated in the glossary, a becak is a pedicab.  Specifically, becaks are a three-wheeler vehicle and commonly used means of transport in Indonesia and other parts of the world.  In most cases, in Aceh, these are motorized in the form of motorcycles with attaching sidecars to seat two passengers.   96  A combination of interview data from various respondents indicated that approximately 75% of households had a motorbike and around 3% of households owned a car.  Most of those who have an occupation travel to the City Centre for work and must pay 10,000 Rupiah29 by public transit each way or travel by scooter for 30 minutes. This is significant given their average earnings of 50,000 to 70,000 per day, based on an average of incomes estimated by block chiefs.  Though there is the coastal area of Neheun nearby, the majority of fishermen in the village still have to go to the main port in the City Centre as they are restricted to membership in that particular fishing union. There are some workshops and brick factories nearby, although the number of brick factories is half of what existed before the tsunami. A detailed and current (2011) map of Neheun Compound was not available; the compound chief indicated that one did not exist.  When consulted, the Neheun village chief and secretary provided an outdated BRR Masterplan map and not all of the compounds shown on the map were built.   When interviewing the youth chief for Neheun Compound at his home, he referred to an aerial photograph when indicating a particular block. He offered a photocopy of the photograph to be used for the research, which is similar to Figure 4.5. below.  As indicated on the photograph, Neheun Compound is organized into seven blocks, A to G, with a total of 606 houses in the completed compound.  Each block is spatially quite distinct and consists of an isolated sub-community.  Block letters increase in conjunction with elevation gain; in other words, Block A is at the bottom of the compound while Block G is at the highest elevation.                                                     29 Exchange rate at time of study: 1$US Dollars = ~9,925 Indonesian Rupiah (Rps)  97   Figure 4.5. Neheun Compound.  Grass area at base was used for temporary barracks for villagers displaced in surrounding villages. Source: Extracted from presentation shared by BAPPEDA Provinsi (with permission) Resettlement process Based on interviews with community members, the process of resettlement to Neheun Compound was entirely non-participatory.  The village was planned, designed and contracted entirely by NGO-A, with international consultants leading the construction process.  NGO-A also carried out housing design, the quality of which most villagers expressed satisfaction with.  Some of the key informants included some ex-combatants, who were positioned as security officers when the housing was being built as a means of reintegrating them with the community.  For their service, the ex-combatants interviewed stated seven of them were rewarded with a house through approval by the chief of district. As described by informants, the process of securing a house in Neheun Compound was facilitated through BRR, which drew on a database of tsunami-affected beneficiaries waiting for a permanent shelter.  Beneficiaries included those unable to return to previous lands now submerged under water, and renters without home ownership. Through a lottery system, they were allocated houses in one of a number of resettlement sites spread across Banda Aceh and Aceh Besar.  The beneficiaries did not have a choice in which resettlement site they would be assigned. Therefore,  98  community members came from a wide assortment of communities and backgrounds across Banda Aceh and Aceh Besar, and similarly had come from many different barracks.  Each block is an assortment of social classes, with one informant explaining, ?There is a lot of disparity within the block, but we have an acceptance of our position.  When people moved here they were already rich or poor and we understand that? (Male, 50s).  Once occupied for ten years, the beneficiary would get ownership of the property. As an informant explained, ?The selling of houses here is not official, because we don?t have the certificate yet.  But we just call it a transfer of the house. It is forbidden for us to sell for ten years? (Male, 40s).  Many ?transfers? of houses had occurred to date.  Community members explained that homes in the lower blocks went for less than those at higher blocks.  For example, houses in Block A would sell for 50 million Rp, while those in Block G would sell for 100 million Rp.  Many of those living at Neheun Compound recalled the government taking down their temporary shelters dispersed around Banda Aceh or Aceh Besar, without having permanent homes ready for them.  Some noted one to two month gaps between barracks being destroyed and their house being available. One of the block chiefs noted that the first eleven families moved to Neheun Compound from the same barrack in September 2007.  Another block chief recalled that in the second batch, eight families came together from a different barrack in October 2007.  His wife gave light that one of the beneficiaries was able to use his connections with BRR officers to facilitate the possession of a collective number of homes circumventing the lottery system. It appeared that several clusters of families had been granted houses through contacts within the barracks.  Others had obtained a house through support from the village chief of their original village.   Even though the mayor of Aceh Besar attended the opening ceremony of the compound in 2007, when families first moved to Neheun Compound, there was no electricity for weeks after.  The community had to be active to collectively request BRR to turn on the electricity to the compound. The water situation was described as being difficult at the start with many community members reporting no running water for quite some time.  There was little livelihood support available to community members in Neheun Compound. A livelihood program from a local NGO had conducted job training for females in the compound a few years ago.  One woman claimed that, ?They gave us the theory and motivation, but no money or tools? (Female, 30s).   99 Conditions in 2011 With the addition of the new compounds, the population of the village of Neheun is estimated to have increased from 4,000 to 12,000 persons. Key informants at the Neheun village office, who were also original habitants of Neheun and living outside of Neheun Compound, remarked on the new compounds, ?they don?t bother us much, because they are so isolated? (Male, 50s).   The community of Neheun Compound therefore is an entirely new community, though housing around 30 people who are originally from Neheun.  One of them commented, ?I grew up in Neheun for all of my life.  When we returned it was like I had moved to Jakarta or Medan.  There are different cultures and so many different ways of doing things here? (Male, 40s). Neheun Compound is not yet integrated with the broader community of Neheun.  One villager in Neheun Compound commented, ?At first there were a lot more problems, such as dealing with the old community of Neheun.  Now it is getting better but they still see us as newcomers? (Male, 40s). On acceptance, a villager compared Neheun Compound to another new compound in Neheun, ?They still have problems being accepted due to jealousy because they have such nice homes and also because they are so close to the original people of Neheun.  We are far away so it is okay? (Male, 40s).   Livelihood struggles are pervasive across the compound.  Most of the residents are from various villages in Banda Aceh, and continue to travel to the city 15 kms away for employment.  A block chief said, ?The ones who are happy here are the ones with a job.  The ones without a job will die if they stay here? (Male, 50s).  As another interviewee explained, ?Many have lost their jobs because they are far from the city and many people are moving back to the city because of that.  They have no choice? (Male, 60s).  Estimates from community members indicate that around 40% of the compound is without stable employment and that many would eventually sell their house and move back to the city or in with family members in the city. One villager explained that, ?An average salary is 50,000 Rp/day, then we spend 20,000 Rp/day for transport, then 15,000 Rp/day for lunch, and then we come home with 15,000 Rp/day.  That?s not enough to survive? (Male, 30s).   Less than 20% of the females in the compound are working.  One of these women explained; You need to have a really formal job for it to be worth it, living here, like a government officer.  When living in our previous villages we had a flexible schedule.  We would work two days, make food, sell it for a day, and work when possible.  Women here can?t do that anymore (Female, 40s).    100  She described some cases of jealousy between couples, when the wife would be working and gone all day while the husband remained at home.    Because of their distance from the main market, men had taken on the responsibility for picking up household groceries from the city on the way home from work. Women commented that they would only go to the city when they had to buy cosmetics. Women complained of loneliness and boredom, being confined to their homes because of the physical geography of the community.  In one woman?s words:  At least before we could have our husbands coming home for lunch. Not anymore.  Before we contributed and worked, like as a seller, shopkeeper, or by doing cleaning jobs.  It was possible to do this, but now it is not possible.  (Female, 30s) Another explained, ?Our husbands take lunch boxes now.  Even our children are taking lunch box to school and don?t come up the hill, home for lunch? (Female, 30s).  Another mentioned that, ?If we have no vehicle then we just sit at home and pray or cook or sleep.  Only our husbands can go to work. We sit at home all day? (Female, 40s). And another,  We have to find a way to pass the time.  Like cooking, sleeping, arranging the children, sleeping? we are very bored now.  Not many women are chit chatting here? we are all from different backgrounds.  I only meet my husband at night time.  We are separated for a long time now. (Female, 30s) A few women are occasionally hired by neighbours to assist in tasks such as housecleaning.  The general atmosphere of Neheun Compound is quiet and empty, especially during weekdays.  There are few people visible on the streets.  With little public space and the steep nature of the compound, the small gatherings that are visible are on the steps of houses, and are mostly comprised of neighbouring women and small children. Bicycles are banned in the compound due to many accidents; in fact, two children were killed in 2007 due to cycling mishaps on the steep hill.  There is some increased activity over evening and weekends. With no market in Neheun Compound or surrounding compounds, many expressed the wish for a local market.  They indicated that there was not enough competition and population to have one.  In discussions, many interviewees drew contrasts to their previous villages of origin, especially when describing their social networks.  In one key informants' words:   101  Life is difficult here.  In our old traditional villages we would have many generations living there? lots of attachment and all of our family would be there during the day.  Here it is independent.  We are only meeting for meetings like to discuss the Maulid30.  In our traditional villages we would be meeting all day? giving food to each other and taking care of one another.  It will take a long time to build the community here? because of so many newcomers and because many people are coming and going all the time. (Female, 50s)   Several homes in Block G have been sold and are being used as weekend villas for their scenic views. In the words of an informant from the Block, ?the rich come and go on the weekends? (Female, 40s).  There are rumors of some villas being used to accommodate second wives, with the first living in the city, however,  ?? this is really secret. We don?t talk about it? (Female, 50s).   The breakdown of occupied and vacant houses appears in Table 4.3 below; these numbers are based on consultations with block chiefs.   Table 4.3. Neheun Compound housing breakdown  Source: Fieldwork When NGO-A built Neheun Compound, they constructed not only houses, but also key purpose-built infrastructure.  These include a mosque (at the base of the hill), a health clinic (in the middle of the hill), a roofed open area for a marketplace (in the middle of the hill), and three school buildings in a horseshoe formation (at the top of the hill).  As the main meeting place in the compound, the mosque is most frequented by male members; social protocol meant that females                                                    30 As translated in the glossary, a Maulid is a celebration of the birth of the Prophet. The occasion falls in the month of Rabi?al-awwal in the Islamic lunar calendar.  As one of the most important official holidays in Indonesia, each village hosts their own lively Maulid celebration during the month.  The celebration is marked by a number of traditions, including serving food and inviting surrounding villages to join in festivities.  The author was able to attend some of these celebrations during her time in Aceh.    Bl c Occupied Unoccupied Total Occupancy RateA 79 19 98 81%B 50 36 86 58%C 20 12 32 63%D 40 60 100 40%E 43 15 58 74%F 60 58 118 51%G 60 54 114 53%Total 352 254 606 58% 102  were not mandated and/or expected to join in daily prayers in the mosque.  While used for daily prayers and community gatherings, Friday prayers are not held at the village mosque, as very few villagers are able to return from the city in time.  The marketplace is abandoned, aside from cattle trying to escape the summer heat.  Though the health clinic and school infrastructure were in place since the opening of the village, the relevant administrative and operational elements within them were not organized until much later.  The health clinic opened in November 2008 and the school opened in late 2009.   There is a young midwife who works and lives in the small health clinic and provides basic services.  She provided insights on the general health conditions in Neheun Compound; the most commonly treatments were for cough, fever, diarrhea, hypertension, first aid, pregnancy and deliveries. There are two known individuals with STDs: one with syphilis and the other with HIV.  The most general condition in the village is skin conditions caused by the poor quality of compound water.  On average the midwife sees five to seven patients a day; the number has gone down from previously being around 15 a day.  There have been no maternal deaths, and two known infant deaths that were because a traditional midwife outside the health institution was used.  Through discussion with community members it became evident that many villagers preferred to go to a senior midwife in the main village of Neheun at the bottom of the hill. This was due to greater confidence and comfort due to her seniority.  Many others still travel to the City Centre 15 kms away to the full service health centre. Only one of the school buildings is being used for two classes, being conducted simultaneously (grades 1 and 2).  Aside from being occupied by cattle and covered with their patties, the remaining two buildings are vacant and in disrepair.  Electrical fixtures and wiring has been removed and many windows shattered.  The back of one of the school buildings shows recent and significant signs of rock fall from higher up the mountain evidenced through boulders resting at the foot of the building.  One of the teachers explained that: It is hard to have a class at the top building.  If it is raining then there will be mudslides and rock fall and children will slip.  Maybe once there is a third class and we are forced to use another building, we will have to figure out what to do.  (Female, 50s)   With little government attention, the school borrows furniture from other schools and until recently was using mattresses as chairs.  Teachers at the school spoke of parent?s preference to send their children to the lower school at the bottom of the hill, because it was more established and  103  ?complete?.  Some interviewees argued that taking their children down the hill to school in the morning is easier than taking them up the hill to the school above.  Teachers at the school spoke of the problem in teaching when children would move back and forth from the city ?6 months here, 6 months there?.  For those consistently living in the village, there was a full attendance rate. In the current academic year, 2 students had left to go back to the city.  A handful of houses dispersed across the compound had been converted into other facilities and/or for service providers.  These include for example, a preschool, a kindergarten (with 20 students), Principal?s office, mechanic shop, convenience store, and a coffee shop. Several women described the small shops in the compound as ?expensive and not complete? (Female, 30s).   The compound has an independent system for accessing water through three different wells.  Parts of the system rely on electric operated pumps, which pose a problem when the electricity is out.  Many alluded to conflict over access, which varies by block and depends on where the house is in relation to the well and how many people the well serves.  Daily access to water ranges from 30 minutes per day to 8 hours a day depending on block.  However, a key informant explained that water is only a problem if one does not have a plastic water tank in their house.  Another key informant interviewed served in the role of water officer for two blocks sharing a well and was elected by the community with the task of turning on and off the water supply to different blocks during the day. He was one of three water officers for the different wells.  The water from the wells is collected in a holding tank and then directed to different blocks, while some other blocks use the electric system.   With no dumpster or garbage collection facilities in Neheun, many community members resorted to using limited excess land to burn waste. Waste that could not be burned, like baby diapers, were either individually taken to the city, or buried in the ground.  A number of interviewees commented on fights with surrounding villages when they used their dumpsters for garbage disposal.  Two houses share one septic tank, a source of conflict among neighbours when it comes time to allocate the cost to have it emptied.   Understanding village priorities was difficult.  Given the number of compounds in Neheun, the Neheun village chief explained that the village budget was allocated to one village at a time. We can?t satisfy all needs with the budget of the government, so we put priority on different areas at a time.  We do one area at a time to make the money worth it.  If we divided the money equally then no one would get enough money to do anything. (Male, 50s)  104  The upcoming priorities for Neheun Compound were to construct a community hall and to install another well.  They had also submitted a request for garbage pickup to the government of Aceh Besar some time ago but had ?not yet heard back of a decision? (Male, 40s). Situated along significant elevation, the community members of Neheun Compound consider themselves safe from future tsunamis.  Few expressed concern over rock fall.  ?It is the most likely risk to us, but it hasn?t happened yet? (Male, 30s).  4.2.2. The compound of Panteriek Overview: a non-participatory case of resettlement not in previous location The compound of Panteriek is categorized as another ?moved, non-participatory? case.  The entire village of Panteriek itself is not a resettlement case, but rather one that contains a new resettlement site.  For this reason, throughout this study the traditional village will be described as "Old Panteriek", while the resettlement will be described as "Panteriek Compound".  The majority of interviewees from both "Old Panteriek" and "Panteriek Compound" alluded to both communities as being two separate communities, the old community (lama) and new community (baro).  However, the village office and leadership structure is considered a joint one, with the village chief representing both Panteriek Compound and Old Panteriek. The village chief had been in his position since 2005.  For this reason, the case study looks only at Panteriek Compound.  However, references will be made to ?Old Panteriek? throughout the rest of this study in order to enable contextual and other important understandings.  Old Panteriek is a traditional village in the subdistrict of Krueng Raya, Banda Aceh, approximately 3 km south of the city centre (see Figure 3.3.).  The community of Old Panteriek experienced little direct impact from the earthquake, as they were not hit by the tsunami wave or debris.  However, they were significantly affected indirectly as a new compound (Panteriek Compound) was built inside Panteriek to administratively become one with the village.  Previously a small community, Panteriek is now considered one of the most crowded ones in Banda Aceh.   As many community members of Old Panteriek described in detail, the land that Panteriek Compound now occupies was previously partially a swampland.  Village office staff noted that prior to the tsunami there were talks with the city of Banda Aceh to convert the land into a park.  However, after the tsunami it became more urgent to build the compound.  Figure 4.6 below, derived from  105  satellite imagery, depicts the area pre- and post-resettlement. As evident in the images, a river flows behind the compound, with only a few meters separating the gates of the new compound and the houses of the old village.     Figure 4.6. Aerial image of Panteriek Compound before (2004) and after (2009) Source: Provided by BAPPEDA Banda Aceh during fieldwork (Google Earth images). Scale unknown.  106  Documentation at the village office was quite thorough for both Panteriek Compound and Old Panteriek, dating back to September 2007.  This was available through the Panteriek village office.  Documentation included monthly reports on population, occupation, education, age and religion.  According to the March 2011 report, the population of Panteriek Compound at the time of fieldwork was 3,209.  The population of Old Panteriek was 1,055 people.  All beneficiaries in the new compound are newcomers to the village from all over Banda Aceh (i.e., coastal villages of Lampulo and Uleelee) and surrounding areas. The population is mixed Acehnese Muslim and Chinese Buddhist, though a very small number of the latter.  Occupations across Panteriek Compound include a fairly equal distribution of fishermen, becak drivers/ builders/constructors, traders/sellers, and government officers.  Most of the households in Panteriek Compound have a motorbike, and a significant proportion own cars.  Panteriek Compound is organized in two blocks namely as: Selenga (type of flower) and Jeumpa (type of flower).  Old Panteriek is also organized in two blocks: Kali (river) and Bambu (bamboo). Each block has its own block chief. Resettlement process Based on interviews with community members, the process of resettlement to Panteriek was entirely non-participatory.  The compound was planned and designed entirely by NGO-B.  The same organization also carried out housing design and contracting of builders.  The village staff described the process of engagement as one where NGO-B approached the mayor of Banda Aceh, who approached the community of Old Panteriek.  However, as the village secretary described: We had no choice in the matter.  The land was owned by the Mayor of Banda Aceh.  Besides, the situation was miserable. And of course we all wanted to help our fellow citizens.  It would not be right to object. (Male, 40s, Old Panteriek)   Some informants claimed that half of the land was owned by a private owner, who sold it to BRR.  The village secretary described how the construction started in 2005 and was built very quickly. Families from many different areas of Banda Aceh were allocated houses in Panteriek Compound. The process of allocating houses was complex and went through several modifications over the course of allocating.  A key member of the planning agency provided a detailed account on this process.  Some highlights include the following:  The initial partnership in allocating homes was  107  between BAPPEDA and NGO-B. However, due to contrasting perspectives of how to allocate housing (i.e., BAPPEDA wishing to remain consistent in its housing allocation approach), the partnership was transferred over to one between Banda Aceh City (the Mayor) and NGO-B.  Some key points from this partnership include a decision for half of the houses to be allocated by NGO-B and half to be decided by Banda Aceh City.  This system worked for up to 400 homes.  After the 400 homes were allocated, the partnership was transferred once again; this time to one between BRR and NGO-B.  The final 300 homes were allocated through a BRR lottery system.  The total number of houses in Panteriek Compound after completion was 716.  All of these houses were occupied in 2011. Due to the complexities described above, the resettlement process conducted by NGO-B was in various phases, with 200 of the houses being initially occupied at the end of 2005/beginning of 2006, and the rest over 2007, 2008 and 2009.  Many of the beneficiaries from the first phase were from the Jantho barracks located approximately 40km south of Banda Aceh and based on a list developed independently by NGO-B, prior to a formalized verification process. Community members indicated that the first phase moved in during 2006. This is supported by accounts of the resettlement by community members.  For example,  200 of the homes were allocated by [NGO-B] to whoever they wanted.  They had their own area of work in the response and early recovery, so maybe they referred those people.  The rest of the homes were allocated by BRR and the Mayor? several of them to government officers. (Male, 60s, Old Panteriek)    Some females described the last ?batch? of people moving to Panteriek Compound to have taken place in May and June of 2009 and situated in the East blocks 10 and 11 (Timur 10 and 11).  This was cited to be 40 ? 50 houses.  Some described households in these blocks to be poorer than the other blocks.   There were mixed understandings of land ownership by residents in Panteriek Compound.  The majority opinion was that each home in the compound had a land certificate that was valid for 20 years.  The confusion was on what would happen after this 20 year timeframe. Some claimed that after 20 years, the land certificate would be renewed or ownership would become in their name. Others, particularly community members in the Old Compound believed that after the 20 years they would have to leave.  A few others in the Old Compound had very different understandings of their presence.  For example,   108  The new elections are coming up, and it has to be someone from here.  The certificate for their house in [NGO-B] is only for 10 years.  So we can't have a candidate from the new compound, because they can only use the buildings and house for 10 years.  So based on that, the new compound members are temporary owners.  What if the government takes their home after 10 years? I think this is possible.  They won't live here for a longer time.  Definitely not for their lifetime, like us? There are maybe two options.  The government will take the land and everyone will be relocated, or the government will kick everyone out and give the homes to government officers. (Male, 60s, Old Panteriek) Conditions in 2011 The village office for all of Panteriek is located just outside of Panteriek Compound, within the area that is technically considered Old Panteriek.  The office is open for set hours almost every day of the week, and it was usually the case that someone was physically present in the office.  In addition to diligence in documenting village data as described earlier, the office was quite keen on keeping records and following administrative procedures.  This was observed through participating in specific research procedures requested by the office when gaining permission to conduct study in the village. There are significant tensions prevalent between Old Panteriek and Panteriek Compound.  The complexity of the situation is evident in the following interview excerpt: The opinion of the new community depends on the individual.  Some will say, yes, I am disturbed and some will say, no, I?m not bothered by them?  More people are disturbed though.  We were told that the new community was for prosperity and betterment when we agreed to the compound.  But really, it is two separate communities.  They have a separate Maulid, korban31, death visitations.  We only come together for Friday prayers because the new masjid is there.  We will have to elect a new chief in August and if it will be someone from the new community it will be a big problem.  But there are larger numbers there so it is a possibility? the old community does not want to be led by someone from the new community.  They don't know the culture here.  We will boycott the elections if they submit a candidate.  (Male, 60s, Old Panteriek)   Almost all of the interviewees described limited interaction between community members across the two communities.  For example, in most cases only the chiefs of blocks are invited to weddings across the two distinct communities, and each community had a separate Imam for                                                    31 Korban (also written Qurban) has been translated in the glossary as sacrifice.  In particular, it is the sacrifice of a livestock animal during the occasion of Eid-ul-Adha in the lunar Islamic calendar.  Eid al-Adha is celebrated by Muslims worldwide to honor the willingness of the Prophet Abraham to sacrifice his son Ishmael as an act of submission to God.  At the last moment, God intervened with a Lamb to sacrifice instead.    109  planning funerals.  Most of the discussions with the Old Panteriek community carried similar negative connotations and impacts of the new compound.  For example: ?There is a proverb, that they need to learn, that all newcomers need to adapt to the new community they join; not the other way around? (Female, 30s, Old Panteriek).  Many of the negativities were associated with differences in culture.  For example :  They have a separate culture.  Such as how to face death. The way it is done in Panteriek is seven days of reading Quran - 5 days at home and 2 days in the mosque.  Everyone from the village should come and do this if it was one united village. (Male, 50s, Old Panteriek) While all of the informants from Old Panteriek spoke extensively of the implications of the new compound, those in the new compound were mostly detached from the tensions.  The new compound seemed to keep to themselves and mind their own business to avoid conflict.  For example, there is limited community involvement in the village office.  As one of the leadership explained: The new community lets the old community take leadership.  They are not too worried that their voice is not there.  They really only go to the village office to obtain their ID card.  They don't want to make the situation complicated. (Male, 40s, Panteriek Compound)   A few members (mostly at the village office) did express positive impacts. For example: Before we were a small village, but now we are big, so we are getting more attention from the city, more assistance from the government, and larger budgets from the government?Some of the budget has been used to build the shops which have become a financial asset to us?   Because it is now very crowded here, many political parties are coming here to advocate and involve the community in politics?There has been an increase in human resources, mainly in terms of the number of government officers living in the new compound.  Therefore there are more decision makers in the community.  (Male, 60s, Old Panteriek).   The general atmosphere of Panteriek Compound is very lively, active, and at times crowded.  At all times there were many children visible playing on the streets, in the football field and on the basketball court.  Small gatherings of people are visible across the compound. This included many groups of females sitting around and chatting.  Due to the dense layout of the compound, at times there was congestion of cars, motorbikes and people. There are many small shops dispersed around the compound as part of, or extensions, of houses (for example, Figure 4.7).  These include water refill shops, food stalls, convenient stores, florists, salons, and so on.  Many women were observed running these small businesses.  The quality  110  of housing was expressed by many community members as being moderate.  Several however, complained of the materials used.  Some of the community members in particular blocks (i.e., Block 11 and 12) complained of water seeping from the ground through the tiles and onto their floors.  A large number of houses in Panteriek compound were currently undergoing renovations, with several houses constructing a second story to their NGO home (for example, Figure 4.8)   Figure 4.7. Shop set up outside NGO house in Panteriek Compound Source: Photograph by author (D. Panjwani)  111   Figure 4.8. Renovations to NGO house in Panteriek Compound Source: Photograph by author (D. Panjwani) Before the tsunami there were no schools in the village of Panteriek.  With the construction of the compound came the establishment of two new schools: a primary school (SD) and a junior high school (SMP), both located in the centre of Panteriek Compound.  Both of the structures for these schools were completed by NGO-B and handed over to the government to institute the schools. The SD school was a new institution established in 2006 and a product of three schools merging, according to the SD principal.  As a consequence, the administration and teachers remained largely the same, but the children were now different.  All of the students were currently classified as Muslim, though, ?at the beginning there were Christian and Chinese students because they were not yet settled? but once they settled they moved their children to Buddhist, Christian schools? (Female, 30s, Panteriek Compound).  Since 2006 to 2011, the enrolment at the school had gone from 155 to 252 students.   As indicated by interviewees within the SMP school, the school was an existing institution that was moved to Panteriek compound because of the large number of students in the compound.  Around 80% of the current student body were cited to be residents of Panteriek; the remaining is from surrounding villages.   The interviewees cited early challenges.  For example, ?when the school  112  first opened, there were many issues.  Attendance was not good, students were dealing with family issues like finding a new occupation, and many had no one at home to watch over their homework.  Over time it is getting much better? (Male, 40s, Panteriek Compound).  While the vast majority of students were Muslim, there were less than five students indicated as Christian or Buddhist.  Since 2006, the enrolment at the school had gone from 63 to 261 students.  While one of the SD interviewees spoke of a UNDP emergency evacuation drill at the school in 2009, they claimed that ?there was no other element to the program, just that one drill.  It is supposed to be done again by the teachers, but we are too busy? (Male, 50s, Panteriek Compound).  The SMP interviewee claimed to have, ?no DRR drills or education because this is usually funded by an NGO are there is no NGO doing a program here.  The government doesn?t do this on its own? (Male, 60s, Panteriek Compound).   There is a government funded health clinic in Panteriek Compound with two practicing midwives.  The health clinic is relatively new and an initiative pushed for by the village chief.  The facility currently occupies a building that was initially built as a village hall.  The midwives reported on average 10 to 12 patients a day, though they only work three hours/day.  The most common visits were related to breathing difficulties and coughing.  One suspected a possible relationship between the gypsum used in the houses and the high occurrence of asthma and allergies, though she cited no research or studies linking the two.  At the time of fieldwork, there was only one room in the clinic being occupied, with the facility primarily being used as a medication centre for basic symptoms.  There were plans to expand the facility over the next year to include a 24-hour midwife and a broader range of services. A visit to the larger health facility just outside of Panteriek Compound, but catering to the community, provided few more details. There were plans to move this larger facility to the existing health clinic in the compound in the following few months to year.  One of the midwives at this facility reported three cases of malnutrition in one of the blocks of Panteriek Compound (in the same family), which had led to health problems in the children.  She also reported many cases of coughing attributed to the material of houses, citing that "they have used gypsum material and the air doesn't flow well because of it" (Female, 30s, Panteriek Compound).  She reported eight to nine births per month, two infant mortalities and no maternal mortalities.  There were two known cases of STDs.   At one end of the compound there is a basketball court, gym and small football field. There were almost always children using the court and/or the field.  For the most part, these facilities are  113  used exclusively by Panteriek Compound although other groups (for example, a nearby university) rent out the gym every so often.    Panteriek Compound has a sophisticated garbage collection system, with a garbage truck coming to pick up garbage from houses three times a week.  There is also a recycling facility that has been built on land leftover at the back of the compound.  Each house has their own septic tank.  There was consistent and reliable electricity and water from the water company in Panteriek Compound.  Some of the lower income households had tried to make a well in their lot, but realized it would give them ?yellow? swamp water.    4.2.3. The village of Gampong Baro Overview: a participatory case of resettlement not in previous location The village of Gampong Baro is a ?moved, participatory? case. Located in the Masjid Raya subdistrict of Aceh Besar, the village is situated right off the main road on a slightly elevated hillside approximately three kilometres from the coastline (see Figure 3.4).  The land that the village now administratively owns and occupies was previously forested area partially belonging to the neighboring village of Durung and partially privately owned by an Imam in the nearby village of Lamnga.  Gampong Baro is the smallest case study explored.  The new site is exactly two hectares in area and is in close proximity to the surrounding village of Durung.  There is a large wired fence surrounding Gampong Baro that specifies its strict borders.    Before the tsunami, the original village of Gampong Baro was located directly on the coastline, around 5km to the West from the current location. The original village occupied around 420 hectares of land.  Due to tsunami impacts, the previous location is now largely submerged in water, with few landmasses remaining and pockets of swampland. All homes were completely destroyed by the tsunami wave.  Though uninhabited at the time of research, administratively, the previous land also belongs to Gampong Baro.  Exact pre-tsunami village data was unavailable.  As the village secretary explained, the habit of recording village data had only come in place in the last few years, due to the influence of international aid. Information from community interviews indicated that over half of the population of Gampong Baro was wiped away by the tsunami.  Estimates allude to 120 survivors and approximately 100 to 150 lives lost. A commemoration plate listing the victims of the tsunami has  114  been placed outside the village office, listing a total of 121 names. The list includes 41 males and 80 females. Some insisted that the list on the monument was accurate, while others felt it was incomplete.  There is a significant inequality evident in the type of victims, with the vast majority being women and young children.  Several informants claimed that of the survivors, only 14 were women and 10 were children aged 5 to 12.  Prior to the tsunami, women claimed there were more than 100 women in the village and large numbers of children. According to village data at the time of research, the village had a current total population of 57 families consisting of a total of 175 people, including 106 male and 69 females. Of the current village population, 74 are children.  The population increase is attributed to new marriages and births. A traditional village, many village members are related and the entire community is Muslim. Prior to the tsunami, Gampong Baro was predominantly a fishing village.  As the village chief explained, ?Before everyone?s profession would automatically be a fisherman, because they were near the coast.  But now they have to take up new occupations like sellers?.  According to the village secretary, of the 43 working members of the village the ?official? occupation breakdown was: 15 fishermen, 13 merchants, 6 fish farmers, 5 government officers, 3 becak drivers and 1 other.  Every day a group of 20 go by motorbike to the coast to catch fish. None of the women have official occupations, though several contribute to livelihoods through selling homemade goods like cakes and occasionally tailoring clothing.   Many villagers own a motorbike through soft loans provided by NGOs.  None of the villagers own a car.  The few land masses remaining of the old land are separated from the mainland by water. There is half a bridge that was commenced as a construction by BRR after the tsunami (see Figure 4.9 below).  However, as villagers explained, the project was stopped after half the bridge was complete. "We ourselves wonder why BRR only built half the bridge" (Male, 30s).  As a consequence, the only way to get to the old land is by a raft that has been constructed by the community through one of their livelihood grants.  Groups of men use the raft to get to the old land in order to conduct livelihood activities (see Figure 4.10 below).    115   Figure 4.9. Half constructed bridge from mainland to old Gampong Baro location Source: Photograph by author (D. Panjwani)  Figure 4.10. Raft used by villagers of Gampong Baro to access old location Source: Photograph by author (D. Panjwani)  116 Resettlement process Based on interviews with community members, the process of resettlement to the current location of Gampong Baro involved several key actors. The village chief at the time did not survive the tsunami. However, a highly respected village elder (i.e., a descendent of noble blood) stepped in as an acting village chief.  Because of his noble title, his position of respect extends across the region and as such, he is not able to officially act as village chief;  however, it was not until 2006 that the community voted for a permanent chief.  The current village secretary, who was in the same position at the time of the disaster, also played a key leadership role in the resettlement process.  Other key actors in the resettlement process included NGO-C, NGO-D and NGO-E.  The resettlement process involved a series of events. Soon after the tsunami, surviving community members of Gampong Baro were sheltered in a school building functioning as a temporary barrack.  The barrack was located on government land across the main road from the land that is now the new village. While staying at the barracks, the community of Gampong Baro was approached by NGO-C offering to provide assistance to build a house.  As explained by a key informant: They said, we want to help you with making a house.  We showed them our land and they said this is wetland and we cannot build here.  After some talk they said if you can get the land we will build houses for you. (Male, 60s)  The community actively sought land through NGO support: [NGO-E] and [NGO-F] had a health clinic down the road.  [NGO-E] was helping the clinic.  When they finished the project they came to us to say goodbye.  We sat together in the barracks with them and when we were talking, we explained our situation, and they asked us what we needed.  We told them that there is [NGO-C] that wants to build us houses but we do not have the land to build them on.  [NGO-E] told us that they could not buy us the land, but they asked our community to make a proposal to them for economic assistance for things like livestock, and making and selling food. We knew the money would be used to buy this land, even though in the proposal we wrote livelihood ? we did not have to submit a report to [NGO-E] showing how the money was used. (Male, 60s)  Villagers explained that the proposal was approved and each of the 57 families received 8 million 700 hundred thousand Rps for a total of 500 million Rps.  Prior to purchasing the land, which was partially owned by the neighboring village of Durung, the village wanted to ensure that they would maintain their own name and administrative authority.  At first, the village of Durung refused and claimed that the land would remain the authority of Durung, even if the new village settled there.   117  However, as explained by the acting chief, through submitting a proposal to maintain its authority, and after much persuasion, Gampong gained the support of leaders including the Regent and Sub-District Head.  With their support and alongside key leaders from Gampong Baro, Durung was approached again.  This time their case was successful.  Through an MOU signed with Durung, the 2 hectares of land were transferred to the authority of Gampong Baro.     With the 500 million Rps attained from the proposal, the community was able to purchase 2 hectares of land. The transaction was done on behalf of the community by the acting village chief. A Memorandum of Understanding was signed between the community and NGO-C prior to the houses being constructed.    [NGO-D] came and normalized the land because before it was all forestland.  We requested this from them because they had a program nearby ? their greenzone program.  Our food assistance was from them, so we went back to them to request help for the normalization.  (Male, 40s) Concurrently, the community had proposed plans to make the government land on which the temporary barracks were situated into permanent housing and land for Gampong Baro. The government owned area covered 400 hectares of land.  A village elder explained, ?We made a proposal to the governor to keep that land, but the process was taking so long. So we kept that ongoing but at the same time we took the initiative to get a donor to buy us land? (Male, 60s).  The community did not hear back on the proposal and over time dismissed the option. The resettlement process that was followed and described was partially participatory.  The current village chief described it to be an ?interactive participatory process?, with a back and forth communication between the NGO-C and the community.  Other elements of participation included that some community members were supervisors during the construction and were responsible for working with the contractors commissioned by NGO-C.  According to some informants, NGO-C had some builders hired themselves, but also paid local village members to help build houses. NGO-D was involved with normalizing the land (which was described as forest).  Village members were also actively involved with normalizing the land prior to building houses. The distribution of houses was not entirely based on land ownership, but rather largely based on family circumstance.  In other words, each married couple acquired 1 house. As a block chief explained, ?If the parents were killed, but 3 children living, and all 3 children got married, then that family would get 3 homes? (Male, 40s). Regardless of how much land a family owned on the  118  previous village land, they were given the same amount of land per couple in the new village.  The community collectively decided that housing would be allocated through a lottery system.  ?We had a lucky draw to assign the houses so that we could minimize jealousy.  Of course we would all want the house that is closest to the main road? (Female, 30s). NGO-C was also actively involved with livelihood assistance with the community during the reconstruction process.  For example, they ?helped teach us how to dry baby fish and gave us a net for catching the fish. They also gave the women sewing machines ? some are still using them? (Male, 50s). In addition, the community of Gampong Baro was engaged early on with a number of livelihood support activities on their old land. These included a number of activities, referred to as the Green Coast Project.  The project was implemented by a group of NGOs with a mutual aim of rehabilitating the coastal environment while nurturing economic empowerment within the community.  At the same time the NGOs sought to develop facilities for environmental education. A key player with the Green Coast project spoke of the willingness to cooperate among leadership and community members: ?At first there were many NGOs coming to us to do work.  They knew that Gampong Baro was a participatory community that was willing to cooperate?. (Male, 50s).  Based on project documents and key informants, activities commenced in August of 2007 with the formation of three groups consisting of 30 villagers and rehabilitation work consisting of the planting of mangrove seedlings and beach vegetation seedlings.  Some projects, such as pine and mangrove planting were completed early on and now are simply maintained and watched over by the community. Decisions to proceed with NGO funded projects such as these were done as a community.  The community also expressed high confidence in the chief as having the best interest of the community in mind during these decisions.   Before we choose to agree to this program we had another offer from the Ministry of Agriculture, but we refused it because the ministry said that they would use their own manpower.  The Chief did not agree with this, that?s why we refused it.  (Male, 60s) The overall recovery process also included a number of small scale activities led by different NGOs.  For example, these include a well, water supply, capacity building, and the provision of items such as computers, bicycles, and non-food items.  The agriculture department had also provided garden rehabilitation.  119 Conditions in 2011 Many of the interviewees made reference to pre-tsunami life as being more favourable than their current situation.    Before the tsunami it was a better time for us.  We had livestock, land, and gardens and we were near the sea.  There was no need for worry and great effort. For example, we had livestock and shrimp ponds.  But now we don?t have these small things in our life.  We are still waiting for government assistance to make shrimp ponds.  (Male, 50s) Yes, we have a better concrete house now, but our financial situation here is hard.  Our source of life is there [previous land].  We would prefer to be back there. Everyone in the village would prefer to live there instead of living here, but unfortunately there are no NGOs here anymore to assist us in this.  (Male, 40s) A fairly compact community, it is mainly the women who are visible in the small common areas that have been created from wood panels.  The women spoke of the greatest impacts of the movement to include less time with their husbands and limited involvement in any livelihood activities.  Other impacts extended to limited social support and personal choices.  For example, ?Before we had a lot of relatives to take care of our children.  Now we have to be careful and have to plan if we want children.  We have to use contraception methods now? (Female, 30s).  All of the homes in Gampong Baro are occupied; many community members complained of the houses being far too small.  Each home is constructed by basic BRR guidelines, measuring 36 square meters and including 2 bedrooms. Within the fenced in land boundaries available to the village, there is no room for the construction of more houses (Figure 4.11 and 4.12).   The amount of houses is not equal to the amount of demand now.  My daughter has married since moving here and is living with her husband and her children in our small house.  We don?t have the land, the occupation, or the money to get a new home or more area. (Male, 60s)  120   Figure 4.11. View from top of new resettlement of Gampong Baro Source: Photograph by author (D. Panjwani)  Figure 4.12. New resettlement of Gampong Baro Source: Photograph by author (D. Panjwani)  121  Since 2007 the villagers have developed other economic activities.  A number of rehabilitation activities on previous Gampong Baro land continue, with new projects being explored through local partners. These include agriculture, selling fish, and animal husbandry of cows, goats and chickens.  For example, the Green Coast recently added a fishpond initiative and dragon fruit plantation; at the time of the research 2 to 3 months had passed since implementation.  Around 15 villagers preserve the land on the old village site, while another 10 are working on other projects on the old land.  These villagers were selected through community meetings.   Female community members are aware of potential livelihood assistance programs through loans, though the majority seemed hesitant to submit proposals.  As one explained, ?We are so afraid that we will not be able to return the money because we are not sure of the future plan? (Female, 20s).  Four small shops have been set up in the village and are being managed by women. There is a deep well at the bottom of the village from which the entire village gets water. One of the youth members of the community manages the well, with a matrix on usage by house.  A water machine controls the distribution of water; on occasions when the machine breaks, villagers travel by motorbike to get water from the city.  On their previous land, each house had their own well that villagers had dug up themselves.  Several of the female villagers complained of the stones beneath the ground preventing them from doing the same on the current land.    There is no school in the village.  Children attend schools in the nearby village of Neheun with fair ease, travelling by public transit or walking.  There is a small health clinic.  With no dumpster or garbage collection facilities in Gampong Baro, all of the villagers burn their garbage themselves on their property.  Each house has its own septic tank.    4.2.4. The village of Bitai Overview: a non-participatory case of resettlement in previous location Bitai is a 'not moved, not participatory' case.  Located in Banda Aceh within the Jaya Baro subdistrict, Bitai is situated approximately 3 km from the ocean, just missing the initial 2 km 'buffer/evacuation zone' (see Figure 3.3).  Bitai was historically on the territory of Aceh Besar and only later did it become an extension of Banda Aceh.  Previously known as a centre of religious studies, it was home to many Islamic teachers and was the place where the Sultan of Aceh (Iskander Muda) learned about Islam.  Consistently cited as a deeply religious village, Bitai continues to have a strong reputation of religious diligence. For example, the village continues to house two Islamic  122  boarding schools and all of the residents of Bitai are officially listed as being Muslim.  However, several informants noted a decrease in their religious attendance and noted the presence of very few Islamic teachers remaining in the current day.    The village of Bitai is one of the sites that holds great significance in the history of Aceh through its relationship with Turkey, dating back to the Ottoman era.  For example, Bitai is named after Teungku Di Bitay from Syria and houses a Turkish graveyard.  It is for this reason that housing in Bitai was completed by NGO-G (associated with the Turkish government).  However, this relationship had been maintained prior to the tsunami, as one of the key informants described: Even before the tsunami we had a good relationship with Turkey.  Every year we would have one or two buses of Turkish coming here to see the cemetery of the son of the Prince.  You can go see the stone in the masjid with his story.  It is because of this that after the tsunami, [NGO-G] are coming directly to our village to help (Male, 30s) One key informant cited figures of around approximately 300 families and up to 2,000 people residing in Bitai prior to the tsunami event. Only 300 people were said to have survived the tsunami.  Current day community data was just starting to be collected, organized and documented systematically.  While some of this data was available from village officer leaders, the majority was obtained from key informants.  According to printouts provided by the village secretary, the current population of Bitai was recorded at 244 families and 820 people, with the majority being young adults (25-40 years old) and young children.  However, the village chief indicated that population numbers may not be entirely accurate due to the constant movement of people in and out of the village.  Reasons for this movement were attributed to the high number of renters in the community.  There were only two original residents over the age of 60, as the author was informed when meeting both of these women at the mosque over the course of the fieldwork.  This number was supported by other villagers, such as the village midwife. Prior to the tsunami, a large percentage of the village population was described as fishermen, farmers and traders. Current occupational data from the village secretary and village leader varied; some of these breakdowns were cited as not being representative of the population due to the high number of renters (i.e., many of the students and government officers).  For example, informants also spoke of a large percentage of individuals claiming residence in Bitai, but in reality residing elsewhere and renting out their property in Bitai (i.e., many of the government officers, police, and business owners). Village office data indicated widespread occupation including: midwives, government officers, teachers, private businesses, students, police, laborers, retirees, fishermen,  123  farmers, carpenters, mechanics, maids, and jobless. The highest numbers consistently fell within private businesses, jobless and students.   The village chief cited 25% of the population as government officers, 5% military officers and 70% as traders, workers and/or jobless.  A large portion of those cited as jobless were original villagers of Bitai.  One key informant cited that less than 20% of original youth are educated at the university level. Resettlement process The tsunami claimed the life of the village chief of Bitai; consequently, the secretary took over the role of  chief.  Though he played this role for two years, the current chief of Bitai had been elected by the community in 2007. Interviews with both the acting chief, current chief and other village leaders provided detailed insights into the resettlement process.  Immediately after the tsunami, many villagers of Bitai were staying at the same barracks in Mataie, in the south of Banda Aceh.  From one month after the tsunami and onwards many of them would visit Bitai to see the remainder of their homes and talk among each other.  Because all existing community data had been lost, the community collectively developed a list of community members, highlighting survivors as they became aware of their whereabouts.  In addition, community members used old foundations of homes to verify the numbers of homes. The acting chief played a key role in compiling data on survivors, which was eventually passed on to the NGO-G government to use in developing housing plans.   The initial contact with NGO-G representatives was described to be in early 2005 (around April), when they visited Bitai and requested community data on survivors.  The acting chief recounted that it was in August that delegates from NGO-G actually came to see what was left of the village and explained to the acting chief that they were keen to build houses and facilities in their village.  In November they built a sample house to show the community what it would look like.  By the time the sample house was built, the community data had already been submitted to NGO-G.  As one leader explained: Many were coming to ask for a house after the sample house was built, but by then the data was closed.  That is why some don't have a [NGO-G] house.  When the sample was built, everyone was coming to get houses - like three members per household - some relatives are also coming to ask for a house for their nephew and so on (Male, 40s).  A key informant explained how NGO-G had a quota of houses that they were able to offer; the quota was greater than the number of houses needed in Bitai.  With only 238 old houses verified  124  in Bitai, the boundaries of Bitai were extended on a map in order to include the neighbouring village of Lampoh Daya. In his words, "We thought, why don't we help our neighbour" (Male, 40s).  As a consequence, 120 NGO-G homes were built in the neighbouring village.  Some village members argued that this was a mistake on the map, and that more houses should have been built in Bitai, which was a source of conflict among community members. However, others claimed that the surrounding village had the same history as Bitai and that historically the original village extended further, which is why NGO-G built there as well. Over the course of the fieldwork, it was common for the original community members to refer to the resettlement process as one that brought up many sensitivities for one or both of the following reasons: not being accounted for by the chief when compiling the list, and/or unhappy that the remainder of homes were allocated to surrounding villages when there was still a need in Bitai.  For example, as an elderly woman who did not get a house explained,  I went to live in my relative?s house for one and a half years right after the tsunami.  When I came back there were no more houses from [NGO-G] for me.  I got a BRR one.  But I accepted it without struggle.  Whatever God gives us, we are content with.  Maybe the chief didn't know I was still alive (Female, Bitai, 60s).  The resettlement process in Bitai has been classified as largely non-participatory, though there are elements of the process that were participatory, demonstrating overlaps in participation and non-participation. For example, there was exchange between contractors and the village leaders, namely the acting chief.  There was also ongoing interaction between villagers and builders, including input in selecting housing design and a back and forth with builders.  For example: We can't really do anything about the model? there was a fixed model.  But we watched the construction and gave suggestions to them.  Sometimes we would buy coffee for the constructor and give them suggestions, like the brick needs water to be secure? little tips to make sure the house was built properly.  Coffee would make them happy and listen to us! (Male, 20s).  Some of the builders and constructors living in Bitai commented how they were not able to get involved in the construction, but simply accepted the house.  However, as described a community member, "to reduce jealousy", the contractors (from Yogya) gave ten houses to the community to build themselves. There were also mechanisms in place between the 'Turkish supervisors' and home owners.  125  We as owners could tell the supervisor if the house was not built right.  The[NGO-G] supervisor would come with his hammer and break the house if it was not built properly? but only two cases like this (Male, 30s).   Several community members said that even though the houses were completed in 2006, many villagers did not want to come back to Bitai until two years later, when they were able to confront their trauma.  The remaining houses were constructed by BRR much later, with the majority being completed by 2008.  This process was also done through the acting chief who requested 177 houses from BRR. Along with housing construction was some degree of land consolidation, whereby the acting chief discussed with community members the option of making roads wider, or creating roads where they did not previously exist.  This was mostly successful with small roads leading to houses, and not so successful for the larger common roads.  For example, "There was discussion to make this road straight instead of curvy, and neighbours would have to give some land in order to do this.  But not many agreed to this" (Male, 40s).   Several other NGOs were involved in Bitai during the recovery process. Activities completed by these NGOs included cash for work schemes, building a school and health clinic, the provision of school equipment and livestock, safety drills, and so on.  There was mention of a livelihood program focused on youth training early in the recovery, though no mention of current livelihood programs in the village. Conditions in 2011 The exact number of houses in Bitai is unconfirmed.  Based on inquiry with block chiefs, the number is around 245.  Based on the village chief, the number is closer to 280.  Based on one village office staff, the number was 324, claiming that 250 were NGO-G home and 74 were BRR homes.  Based on another village staff, 227 of the houses in Bitai were constructed by the NGO-G, while the remaining 53 houses were done by BRR.  Regardless of the accurate figure, the current number of houses in Bitai exceeds the previous number of houses, which was reported to be around 200.  As one community member explained, "Everyone got a house, some broke up their family card and asked for multiple houses.  If it was up to the community to request houses, they would have requested one per family member!" (Male, 30s).  All, but one of the village Block Chiefs claimed 100% occupancy.  The one exception claimed that there were five empty houses in his block.  The quality of houses constructed by NGO-G was consistently cited across Bitai and Banda Aceh as  126  among the best constructed in the aftermath of the tsunami (Figure 4.13).  Some renovations were observed (Figure 4.14).   Figure 4.13. NGO house in Bitai Source: Photograph by author (D. Panjwani)  Figure 4.14. Renovation to NGO house in Bitai Source: Photograph by author (D. Panjwani)  127  Even though there is a building for the village office, it was not in use.  As the chief explained, "because all of the staff in the desa [village] office have other occupations, I opened the office in my house. It is more convenient this way" (Male, 40s).  This was new to the village of Bitai.  As an elderly woman explained,  Even the [leader] is too busy with his government office job now.  Before, the leaders wouldn't have another job and he had time for the community.  Before, the chief knew everything about culture, but now the modern chief doesn't really know.  For example, before if there was a marriage proposal, the chief would go along.  If there was a birth, he would come to name the baby. There is none of that now.  (Female, 60s).  The leadership staff is significantly younger than the norm across the other cases, with the average age appearing to be in the mid-30s.  As one key informant cited: "It's funny when we are having a community meeting now.  It is supposed to be the elderly leading it, but now it is young members leading because there are only one or two elderly left" (Female, 30s).  There is limited transparency from the village office in respect to budgets and financial reporting, in the absence of a designated village office.  There is a health clinic in Bitai, with a fulltime midwife and basic medical equipment including a delivery bed. She described the average number of patients to be five to seven a day; most being children and babies.  The most common illnesses were coughing and/or fever.  There were no reported STD's.    She cited that on average there were one to two babies being born every month in Bitai, with most mothers being between the ages of 20 to 30; there was one case of infant mortality in the last year.  Overall, she described the community to be quite conscious and aware of their heath.  In terms of education facilities, there is a preschool, a kindergarten, and a SD school in Bitai. Like health, education is regarded highly by the community.  The preschool opened in 2009 and had 35 students enrolled and 4 teachers, at the time of data collection. As one of the teachers cited, the population of students had more than doubled since 2009.  She noted a 5 student increase each year, starting with 15 students in 2009, attributing the increase to new marriages of survivors and births after the tsunami. The preschool was in a very crowded space occupying half of the health clinic building.  Originally, the entire building was meant to be the health clinic.   At the time of fieldwork, the kindergarten was occupying a structure that was initially constructed with the intention of being a meeting hall for the village. According to kindergarten teachers, the structure had not yet been used for purposes beyond their classes.  The reason for  128  occupying the structure was because the new health clinic had taken the land where the kindergarten used to be. As a teacher explained: Maybe they forgot that there was a kindergarten here before.  We have been moving around since 2005.  In 2006, we rented a house, in 2007 and 08 we had a temporary shelter, in 2009 we were in the primary school building, in 2010 in the PKK building, and now, since 2011 we are here in the meeting hall (Female, 30s).  Kindergarten teachers indicated that only five children survived the tsunami.  This is supported by attendance levels at the kindergarten, which had increased from five children in 2006 to 25 students in 2011. The SD school (elementary) has nine teachers and 50 students ranging from class one to five. When it opened in 2005, there were 26 students. As the principal explained, because the school had opened late, many students had registered at other, more established schools.  Most of the students at this particular school were cited as being "from poor families".  As a teacher explained, "if you are richer, your child won't go here!" (Female, 30s).  The teacher described how many of the children who had a 'new' mother were the ones with more problems at school, in comparison to the ones with a 'new' father - 'new' implying a marriage after the tsunami due to the loss of one partner.  Another teacher described serious challenges with children who adopted new siblings due to re-marriage.   Interviewees at all three of the schools reported similar behaviours by parents in relation to hazards. For example, Attendance is good here - only when there is bad weather, children will disappear.  After the tsunami, parents are more aware of the weather.  If there is hard rain or strong wind, they come and take their children home.  Before, it was not like this.  Maybe this is because the parents have the experience of the tsunami.  But children born after the tsunami don't know what it is so they don't understand why they are going home.  This is especially true when there is an earthquake - parents come quickly and take their kids. All people connect anything to do with nature to the tsunami.  (Female, 40s).  While the SD school had evacuation drills in their curriculum, the kindergarten and preschool did not. During focus group discussions with the original community members, there was significant discussion on the loss of cultural norms and unity.  Most profound of cultural changes included shifts toward a culture of individualism, where people were described as only gathering for formal events. For example two informants alluded to an example of the Maulid celebration:   129  The culture here is changing completely.  Before, for the Maulid we are all cooking in a huge pot together and then displaying the food in wood platters to take to the mosque.  Now, everyone is taking rice box and just dropping them off! (Female, 30s).   There were many cultural implications cited by villagers on the loss of elderly in the community.  For example: Our parents died, and now there are only youth who are left.  And the youth don't have the attention of parents anymore and this has impacts on our current situation. I mean, if we don't want to go to school, no one forces us to go. (Male, 20s).    On average village members claimed the population to be 50% original and 50% newcomers at the current time. Many of the newcomers were said to be from other parts of Aceh (i.e., Pidie, Jaya, Sigli, and so on).  One village leader explained how Bitai had a new reputation of ?being a good place to rent a good, new house?.  Interviewees estimated that 50% of the population was well off, while 50% was not so well off.  Beyond the cultural implications just mentioned, there is little tension between the original and newcomer population.  Many original villagers reported the newcomers simply being busy going to work in the morning and coming home at night. As a few informants described them "They are silent".  Another commented that, "the most obvious thing is that they don't come to ceremonies and village events" (Male, 20s).  Most of the newcomers were reported to be more educated than the original residents of Bitai.   The general motivation among original community members was low, with high prevalence of comments like, "If your whole family is gone, then there are no triggers to do better.  You keep thinking, what for? (Male, 20s) and ?Our spirit is not life before.  When you have lost half your life, now you are only thinking about the next day.   It is useless to think long-term?. (Male, 30s). Activities that were once very strong in Bitai had not been re-established.  For example, prior to the tsunami, there was a well-known volleyball team in the village that youth spoke of.  Many original community members noted the decreased attendance at the masjid (mosque) for prayers, even among the original community members.  For example, "there are less people in the masjid now, and there is less trust in the community" (Male, 40s).   Over half of the key informants in Bitai mentioned how the new housing was disguising poverty, compared to traditional housing (see Figure 4.15 and 4.16).  For example: Now neighbours don't even know if their own neighbour is doing well or not.  And we aren't really collecting any data on whether households are rich or poor.  This is especially the case  130  for the newcomers.  We don't know if they are doing well or not.  Sometimes we realize that a family we thought was rich actually doesn't eat food for full days at a time.  (Female, 40s)  Figure 4.15. Remains of a traditional house in Bitai Source: Photograph by author (D. Panjwani)  Figure 4.16. Traditional wooden house in Bitai  Built independently by community member. Source: Photograph by author (D. Panjwani)  131  It was evident that the community atmosphere in Bitai is quite quiet.  This was supported by community discussions.  For example: Before, we had many elderly in our community, and it was very crowded with people gathering around in shops and just hanging around.  Now it looks like an organized community with hardly anyone visible. (Female, 30s)  For Eid-ul-Fitr32 it will be a very silent village, you will see.  That never happened before.  The newcomers are going to their family homes, and the youth to their relatives? houses out of town because they don?t want to spend it alone.  It is like a dead village during Eid (Male, 30s) There are very few home industries or small businesses visible around the village. Many cited conditions prior to the tsunami as being more favorable, especially in relation to livelihood.  For example: We still had rice fields, fish ponds, and swamp land.  We could get oysters and crabs.  Youth would always do that.  Now there is a big drainage there and oysters and crabs are not there anymore.  The land is changing too.  Before, we can plant chilli, but now a big part of our land is not fertile anymore.  It seems that now we can only have a livelihood if we are doing a proposal? Youth don't know where to begin.  It's not that we don't want to do it   (Male, 30s)   Joblessness is the biggest problem.  Lots of youth lost parents and need guidance on what to do.  There's a problem on how to train them to be independent.  They have no capabilities or skills or guidance to figure things out.  Many of their diplomas and certificates were lost.  They could arrange to get a copy but there is no spirit.  We lost our parents, certificates, and spirit.  Most of the people left in Bitai at the time were around 18 years of age.  (Male, 30s)   Most of the youth have a low level of education and their spirit to develop or progress is the problem.  Youth are not thinking long-term, but are only thinking of grabbing a job opportunity as it comes.  They are thinking of ways to get money instantly, rather than working hard to find a stable job.  (Male, 20s)   Most villagers of Bitai have consistent water from the water company.  However, several complained about water flow not working regularly, in spite of company pipes.  As a consequence, many villagers had their own small wells.  Some commented that the well water was dirty and suitable only for showering, and not for drinking.  Consequently, many bought water from the market                                                    32 Eid-ul-Fitr follows the lunar Islamic calendar and is celebrated by Muslims worldwide to mark the end of Ramadan, the holy month of fasting.    132  for cooking and drinking.  There is a village garbage dumpster, though it was consistently overflowing and not large enough to accommodate the waste.   4.2.5. The village of Lampulo Overview: a participatory case of resettlement in previous location Lampulo is a ?not moved, participatory? case.  Located directly along the coast and in immediate proximity to the city centre, Lampulo is part of the Kuta Alam subdistrict of Banda Aceh (see Figure 3.3). Traditionally known as a fishing village, Lampulo has a long ancestral history and governance history that has been recorded back to the 1930s.  Based on a traditional story recorded by a past chief, the name of the village comes from characteristics of the area. The basis of ?Lam? (in Indonesia meaning drowning) was given due to the high occurrence of the river flooding the lands.  The basis of ?Pulo? (in Indonesia meaning island) was that the area was previously a lush forest.   According to the initial spatial plan developed by BAPPEDA, Lampulo remains situated in what was considered the ?Evacuation zone?. However, the village was re-developed in spite of this.  The landscape of Lampulo contains significant amounts of wetland.  For example, many areas across Lampulo continue to show signs of floods and/or areas of swamp.  Several community members claimed that their land was normalized using sand and soil in order to rebuild their home. While these houses are on solid land, areas surrounding some of these homes are wet, either due to flooding or swamp (Figures 4.17 and 4.18).  Many households continue to harvest their own fish/shrimp ponds on their land.   133   Figure 4.17. Wetland in Lampulo Source: Photograph by author (D. Panjwani)  Figure 4.18. Traditional fish/shrimp pond next to NGO house in Lampulo Source: Photograph by author (D. Panjwani)  134  The Chief of Lampulo at the time of fieldwork had been in the position for three years; at the time of the tsunami he was a block chief in Lampulo.  The village office in Lampulo seems quite sophisticated in the production and archiving of village documentation with detailed records from the tsunami onwards; all data from prior to the tsunami was destroyed. For example, some of the reports attained from the village secretary included monthly matrices on population, births, deaths, marriages, education levels and occupations.  The village chief was computer savvy and had completed higher education; during the time of fieldwork, he was in the process of developing a website for the village of Lampulo.  In addition, the village office had developed some paraphernalia for purposes of marketing Lampulo as a tourist destination (i.e., framed photographs). It was estimated that there were 6,000 people living in Lampulo before the tsunami.  Only 1,500 survived the event.  Village leaders cited that over half of the victims were female (exact numbers were not available).  With the addition of newcomers, the population has increased to around 4,000 people, with around 2,200 males and 1,800 females.  Village documents indicate that 55% of the current population was newcomer, while 45% are original community members of Lampulo.  The majority of the village is Muslim, with a small percentage of Christians (documents indicate 0.21%).  One of the community leaders claimed that prior to the tsunami, 70% of Lampulo were fishermen.  At the time of the fieldwork he claimed that number had dropped to around 40%, with more people having turned to small business initiatives.  These estimates were supported by some of the block chiefs and community members.  Occupational records of the adult population at the village office indicate: 35% fishermen, 25% student, 15% trader, 15% government officer, 6% worker, 3% handyman/carpenter and 1% military and police.   As described by some of the elderly in the community, Lampulo had experienced substantial growth over the decades leading up to the tsunami. Some descriptions that pointed to this growth included the formation of new blocks over the 1980s and 1990s.  These blocks were described as previously being rice fields or areas of forest and shrubbery.  One community member described how the original village of Lampulo was Block 1, with the other 3 blocks being expansions over time.    Lampulo is currently organized in four blocks: Teuku Tuan Dipulo (Block 1), Malahayati (Block 2), Tgk. Disayang (Block 3), and Teuku Teungoh (Block 4); the same blocks were in place prior to the tsunami.  When going through the village, the boundaries of each block are not very distinct.  Each block has their own leader that is elected by the community.   135 Resettlement process Immediately after the tsunami event, community members from Lampulo left the village and went to either reside in tents or with relatives in surrounding areas.  The chief described that it was only three months onwards that most people, including himself returned to see the aid situation.  Community members were situated in many different barracks at this time; as he claimed ?people were not arranged, they would just go wherever could accommodate them? (Male, 50s).  There were some barracks set up in Lampulo as well, where solely members of Lampulo stayed in order to remain close to their land.  Some community members spoke of even remaining in a tent next to the remains of their house instead of the barracks. For example, a woman in one of the blocks described around 20 families opting to live in tents on their land instead of barracks far away.    As one key informant described, ?during the recovery process many people were concerned with defending their land? people were very active in coming from their barracks to here to do this? (Female, 40s).   In the end, several different NGOs built houses in Lampulo.  However, the overwhelming majority of houses in Lampulo were completed by NGO-H.  The NGO-H houses were also the ones consistently cited as the highest quality houses out of those listed, though they were the slowest. For example, some of the other houses built by another NGO had a cracking foundation that was visible.  However, the same houses were built much sooner, with some community members recalling having returned to their home in Lampulo eight months after the tsunami.     According to one of the informants who was a key leader in the housing process, he went to NGO-H himself with community data obtained from the chief to attract them to Lampulo.  The informant also helped in the post-disaster assessment of Lampulo for NGO-H.  Collectively leaders from the village decided to proceed with NGO-H for the majority of the houses being built from this point onwards.  One of these leaders cited reasons for choosing NGO-H to include the fact that ?[NGO-H] would provide other programs, like psychosocial health and livelihood support? not just homes? (Male, 60s).  Those who were not able to get a NGO-H house requested a BRR one much later in the process.  Participation in the housing process was evident at various stages.  For example, as explained by one of the NGO-H supervisors from Lampulo, NGO-H hired supervisors from Lampulo and ?if you had the skill of supervision, you could apply to [NGO-H]? (Female, 30s). The Lampulo supervisor would report to the NGO-H supervisor.   The current chief of Lampulo was one of these supervisors for NGO-H.  Most of the builders, on the other hand, were described as being from  136  Medan.  Community members simply had a say in housing design.  For example, beneficiaries were given the choice of five different styles of homes, from which they would choose one based on preference.  The block chief would facilitate the process of providing beneficiaries with a form to complete.   Community members who were not in a supervisor role would still watch over their own homes being constructed and would report to the Lampulo supervisor with any concerns.   We were not involved in the construction of the home? we can only watch it being built, because it is not very easy to approach the builders? only some people were about to do that, for us it was just easier to wait and watch until the house was ready (Female, 30s). I know how many iron pieces per house, so I can make sure that the same number is used by the builder for our house, and I check with the neighbour?s house on what was done there, and also with the supervisor on what should be done (Female, 40s)  The housing process through NGO-H was described by villagers as being well intentioned and with a strong start, but having many challenges as time progressed.  For example, in one interview with a Lampulo supervisor he described that ?there was not really any good harmony between the one who works in the field and the one in the office? [NGO-H] changed their leadership three times and the participatory process changed with it? and sometimes the data went missing with it? (Male, 60s). Others described it as one that the NGO was not able to keep in their control. As cited by a community leader ?the quality of houses is not good...the money they had was good, but the people working below were very corrupt? (Male, 50s).  Villagers spoke of the numerous challenges in materials during the building process.  For example:  There were many problems in materials and a shortage of materials, for example the door would not be of good material, but was put there so that at least there would be a door.  But the new manager would kick it out because it was not good enough.  They would give the order to demolish and build again, and this made the process slower? and builders would not get payment if the house was demolished.  Also, the builder would have to come to the office to get the payment, but they wouldn?t have a vehicle so it would be a big problem.  Things like this make it difficult and make the builders not want to work anymore. Many houses were demolished and rebuilt, but many others were left for the community to finish themselves as they ran out of time.  (Male, 60s)  Beneficiaries started moving into their new homes from 2007 (first phase) to 2009 (last phase).  Many beneficiaries claimed that their houses remained incomplete or that it was taking too long and they ended up completing portions on their own (i.e., floor and ceiling). For example:   137  We got 6 million Rp for a 6 month rental from [NGO-H] so that we could rent a house while they renovated ours.  We came and moved back after that and nothing had been done? and because the quality was at still not at par we were told to rent out again, but we said it?s okay we don?t want to move over and over again.  I had to sign a letter that [NGO-H] was not responsible if there is a problem with the house, because I just wanted to live here by that point.  We were tired of the different supervisors? stories, we just wanted our house.  (Female, 40s)  There were several other agencies involved in Lampulo.  These included psychosocial health schemes by NGO-H, Cash for Work programs, and livelihood schemes. Livelihood schemes include small business loans through NGOs and BRR initiatives like women?s training for fish processing, fish nuggets and fish chips.  Since Lampulo is home to the fish market, a significant amount of aid was also directed at rehabilitating the fisheries industry (including a number of different NGOs).  This included a fisherman halls, ice factory, boats, pier, and fishing equipment. Conditions in 2011 As reported by many of the interviewees, and is visible from the number of housing lots, the number of houses in Lampulo had increased dramatically.  As one informant stated: The terms and conditions are like this? if you have your own family then you get your own house, even if you didn?t have a house before.  In many lots where there was one house, there is now two or three.  (Male, 50s).  The village office cited 770 housing units before the tsunami (Figure 4.19).  These included 525 permanent units, 175 semi-permanent units and 70 wood units.  They cited the current number to be 1,109 housing units; all were cited by the village office as being permanent.  However, several temporary wooden shelters being used as permanent homes were also observed (Figure 4.20).    138   Figure 4.19. Remains of traditional Lampulo house Source: Photograph by author (D. Panjwani)  Figure 4.20. Temporary shelters being used permanently Source: Photograph by author (D. Panjwani)  139  Due to the ideal location for fishermen, there was high demand for houses in Lampulo, and those that were not occupied were rented out.  Village members claimed that prior to the tsunami houses were occupied by homeowners, while at the time of this study houses were more likely to be occupied by renters.  Unoccupied housing was unheard of prior to the tsunami.  Block chiefs claimed that all of the houses in Blocks 1, 2 and 3 were occupied.  The block chief for Block 4 claimed around 10 houses that were empty and owners had not rented them out yet.  When asked why, he stated that it was ?because of trauma and that people don?t want to be reminded of the death of their mother or family? (Male, 50s).   There were some distinct characteristics and divisions across the community of Lampulo; these existed from prior to the tsunami.  Blocks 1 and 3 consisting primarily of fishermen and of lower income, while Blocks 2 to 4 were more mixed occupations and of moderate to higher income. Block 3 was described by many as a slum area with a lot of illegal housing, prior to the tsunami, though they now had NGO housing. The homes situated further away from the water and closer to the main road (i.e., Blocks 2 and 4) had greater ownership by government officers.  Several of the homes in Block 2 (approximately 100) had experienced less damage from the tsunami and remained standing (Figure 4.21).   Figure 4.21. An NGO house next to a traditional house that survived the tsunami.   As a much larger structure, traditional houses typically housed multiple families.  140  Source: Photograph by author (D. Panjwani) The general atmosphere of Lampulo was quite lively, especially with the activity of the fishing market along the edge of the village. While most men are engaged with their fishing activities, women in Lampulo are quite active in village affairs.  This includes helping each other or helping the village office in various events.  There are several women carrying out home industries.  A key leader of the village woman?s group cited over twenty home industries, like small shops, bakeries and catering services, in Lampulo.  She spoke of these individuals starting to build collective groups now.  Several original residents, including groups of women cited stronger unity since the tsunami. There is a kindergarten and an elementary school in Lampulo. Based on conversations with some teachers at the kindergarten, the school was rebuilt in 2005 through aid from two different NGOs. Prior to the tsunami, teachers cited 90 students enrolled, of which 7 survived the disaster.  The current enrolment was at 40 students, most from new marriages after the tsunami. The teachers reported severe trauma initially among survivors, with some still prevalent.  For example, ?? he is 12 and can?t stand to hear the wind or see clouds now? (Female, 40s).  The elementary school was rebuilt by an international corporation and is branded as such.  As described by a key informant, ?this school was here before the tsunami, but before the condition was not as good as it is now? now it is maybe 100 times better than before!? (Female, 50s). Based on interviewees and conversations with school teachers and the principal, prior to the tsunami the school had around 250 students.  In their first month of reopening they had 7 students.  Enrolment at the time of fieldwork was of 134 students.  Eight of the 15 teachers were lost to the tsunami, though 7 more have been hired since. The elementary school had a Red Cross DRR program at the school which includes a safety drill.  The last drill had been conducted in 2010 and the teacher spoke of a strong commitment to conduct it annually.   The health clinic was built by another international corporation and is branded as such; the clinic was established in 2006.  The facility is quite comprehensive and is equipped with a delivery room, youth clinic, pharmacy, baby room, emergency room and dental clinic. The facility is a busy one with an average of 60 patients a day.  As described by the doctor at the clinic, the most frequent cases treated were fever, stomach ailments, colds, respiratory problems, and hypertension, arthritis and skin infections.  There were a few maternal and infant mortalities reported. The doctor spoke  141  extensively on the dirty environment in Lampulo as having negative implications on health.  For example, the poor drainage and pools of water in Lampulo tended to attract many mosquitoes, contributing to one to two case of dengue per month as a consequence.  There were two reported incidents of STDs, though the doctor alluded to many more that are not classified as such.    Significant changes to the landscape were reported by over half of the community members interviewed.  These include the increase in swamp land and the loss of tobacco plants, palm trees, fish ponds, and mangroves.  Most of the fish ponds and shrimp pools have disappeared, making traditional livelihood options difficult.   Several key informants involved in the resettlement process described there to be limited coordination between the different infrastructures (i.e., drainage, housing and roads) which has resulted in problems in the current day. In particular, clean water, drainage and sanitation were growing problems for Lampulo. Even areas that did not previously have flooding (i.e., Block 2) where now facing challenges with water seeping through tiles.  Some other parts of Lampulo (i.e., Block 4) faced drainage problems due to construction issues:  When it is raining, the water is coming into the house because the road is higher than the floor of the homes.  The house was built by [NGO-H] first and then the road by the government? the house in 2007 and the road in 2009 (Male, 60s). Blocks 2, 3 and 4 have garbage dumpsters and truck pickup every 3 days, but Block 1 does not have either.  Consequently, community members of Block 1 burn their garbage.  Each house has their own septic tank and water tank pipe.  Access to water is not equal across the community, with many parts not being reached by the water company.  However, there is water company water in the majority of areas for those who are paying for it; however, in most cases even water company water was not consistent due to poor piping.  The well water in Lampulo is not usable.   4.2.6. Summary of case descriptions Resettlement  While basic groupings of ?moved? and ?not moved? will stand for the remainder of the dissertation, findings point to a need to understand resettlement on a range of issues. In adopting this understanding, resettlement has been summarized based on a spectrum that incorporates movement, distance, village structure and community makeup.  A breakdown is shown in Table 4.4 below, with the highest ranked (Bitai) indicating the lowest degree of displacement, and the lowest ranked  142  (Neheun Compound) indicating the highest degree of displacement.  These rankings and specific elements of the breakdown (i.e., distance, village structure, community makeup) will be referred to at various points in the chapters that follow.    Table 4.4. Resettlement spectrum Participation   While during case selection, 'participatory' and 'non-participatory' was used, soon into fieldwork, the challenge of defining participation in the context of post-tsunami Aceh emerged (Mahdi, 2009).  It was hard to say what constituted as being participatory and it can be argued that none of the village cases were in fact truly participatory in resettlement activities, and that in many of the cases, participation and non-participation overlapped. Even though the aid community acknowledged the importance of participation and community consultative processes, they often failed in the implementation of these principles and processes (Dercon and Kusumawijaya, 2007).  For example, many NGOs in Aceh adopted participatory approaches, yet had to change elements of the approach due to time constraints and various complications. Therefore, in the summary here, and for purposes of further analysis, the villages will be ranked relative to each other and based on degrees of participation gathered through interview insights. These are displayed in Table 4.5 below.  Rank Village Location Displacement Village Structure Composition1 Bitairemained on place of originN/A (remained urban)village maintaining own village structure  many original community members, and several newcomers2 Lampuloremained on place of originN/A (remained urban)village maintaining own village structure  several original community members, and many newcomers3 Gampong Baromoved to new locationrural to rural (5km from previous) village maintaining own village structure  original community members 4Panteriek Compoundmoved to new locationurban to urbannew compound integrating into existing village structurevillagers from many different villages5Neheun Compoundmoved to new locationurban to rural (15 km from city)new compound integrating into existing village structurevillagers from many different villages 143  Relative rankings will be used as a basis of exploration against other relative rankings (i.e. resettlement and recovery).       The lack of participation of NGOs in this study limits the understanding of the actual participatory process outside of what is displayed in the ranking.  Therefore, it is difficult to compare these rankings against a traditional ?ladder of participation? such as that proposed by Choguill (1996) and Arnstein (1969).  Furthermore, two of the non-participatory cases (Panteriek Compound and Neheun Compound) included community members who were completely detached from the respective NGO until randomly being allocated a home in a particular village.  Therefore, elements of manipulation and neglect (i.e. lowest end of participation on the ladder) do not necessarily apply to the participation rankings in the way that the ladder is developed. However, what was evident through interviews with community members was that there were elements of empowerment (i.e. highest end of participation on the ladder, and described as actual control in the situation) in the resettlement process described by villagers of Gampong Baro and Lampulo, which were completely absent in the other three villages.      Table 4.5. Participation ranking of cases In the cases of Neheun Compound and Panteriek Compound, grouped as non-participatory cases, the villages were constructed on new land that had previously been unoccupied. It was only Rank Village Level of Participation1 Gampong Baroactive in finding a resettlement locationinvolved in developing the proposal to aquire land active in normalizing and clearing the landinvolved in the rebuilding process2 Lampuloinvolved in choosing an NGO to assist in housinginvolved in selecting housing design, but not the rebuilding processsome village members acted as supervisors3 Bitaiinvolved with selecting housing design had limited back and forth with builders4 Neheun Compoundlocal community members stationed as security guards during the construction process5 Panteriek Compoundno level of communication with constructors until after the resettlement was complete 144  after the housing and basic infrastructure was completed that community members were allocated houses that they would occupy. Therefore community members had no voice in the construction or allocation process.  Neheun Compound has been ranked higher than Panteriek for the reason that local community members were stationed as security guards during the construction process, as described in the village description earlier.  During interviews with some of these individuals, who now have homes in the village, the villagers expressed some level of communication between the constructors and themselves.  Panteriek, however, is perhaps a truly non-participatory case as villagers had nearly no level of communication with constructors or village authorities until after the resettlement was complete.  Bitai, Lampulo, and Gampong Baro have been grouped as participatory cases.  As described in their descriptions earlier, Gampong Baro was the highest case of participation as the community was active in finding a resettlement location, developing the proposal and in the rebuilding process. Lampulo has been ranked as the second highest case as villagers were involved in engaging the NGO, the village planning and in selecting housing design, but not the rebuilding process.  Bitai is ranked as a third case as villagers were simply involved with selecting housing design and had limited back and forth with builders. These rankings are further supported by the percentage of key informants and focus group participants who responded favorably when asked if they were active in the resettlement process. Key descriptors Key descriptors that have been discussed across the five cases are summarized in Table 4.6 below.  These descriptors are important in enabling an understanding of the case studies, even though historical and political contextual underpinnings are not explored in substantial depth.  Some of the data that has been presented as descriptors will be used in the next chapter to develop an assessment of community recovery.  This assessment will ultimately lead to a recovery ranking of the five cases that have just been described.  The analysis will then return to the resettlement and participation rankings summarized above to draw basic correlations with the recovery outcomes.    145   Table 4.6. Key descriptors of cases  Case StudyPopulation before disasterOriginal Population after disaster (Survivors)Total Population after disaster (Original + Newcomers)Traditional livelihoods Significant completion of rebuildingLeadership StyleNeheun CompoundN/A N/A 1,000FishingTrade2007Perscriptive, minimal daily involvement with communityBitai 2,000 300 820Farming FishingTrade2006Detached, minimal daily involvement with communityGampong Baro241 120 175FarmingFishing Trade2006Inclusive, siginificant daily involvment with communityLampulo 6,000 1,500 4,000FishingGovernment2008Perscriptive, significant daily involvement with communityPanteriek CompoundN/A N/A 3,209Fishing Government2006Inclusive, siginificant daily involvment with community 146  CHAPTER FIVE: ASSESSING COMMUNITY RECOVERY Chapter Five aims to answer the first of the four research sub-questions of this study in the context of the five case communities: ?How can long-term holistic community disaster recovery be systematically assessed?? and begins to explore the second sub-question: ?What are some of the key factors that influence long-term holistic community disaster recovery at the village level??  In addressing these questions, the chapter provides an assessment and ranking of community recovery, our dependent variable, across cases. Recovery indicators used in this assessment draw largely from details presented in Chapter Four, in addition to other indicator data that have not yet been presented, but were collected using methods described in Chapter Three.  As described in the next sections, recovery data were organized through a series of indicators, and analyzed through the process of an index that incorporated these indicators.  Once the data were incorporated into the index, comparisons across cases were conducted. Details on the development of the index and calculations are addressed within this chapter and further described in Appendices One and Two. An analysis of findings sets up the base for exploring subsequent research questions in Chapter Six.  5.1. Background In this study, a capabilities-based approach is operationalized in the assessment of recovery in order to enable a multi-dimensional understanding of holistic recovery within the context of community development. As described in Chapter Two, there are limited applications of a capabilities approach to disaster assessment. One of these applications described in Chapter Two include Gardoni and Murphy?s (2008) Disaster Impact Index (DII).  Some guidance from their implementation includes the following: capabilities criteria (ibid. p.326) should consider their relevance, importance, influenceability and practical implementability and accuracy; indicators criteria (ibid. p.327) should be representative of the corresponding capability and be intuitively plausible.  5.1.1. Implementation of a capabilities-based approach An assessment of community recovery is conducted through developing and using a multidimensional recovery index (MDRI) that operationalizes the capability approach beyond a general welfare framework.  As capabilities are not directly quantifiable, indicators are used to measure each of the selected capabilities.   Community indicators that are developed are based on  147  capabilities identified during fieldwork as being relevant to recovery within the context of post-tsunami Aceh.  The process relies largely on transforming qualitative, interview and observational primary data into multiple indicators grouped within selected capabilities.  Indicators are converted into a uniform scale to create multiple capability sub-indices; capability sub-indices are then totalled to create the recovery index. As a composite measure, the recovery index will allow for an assessment of recovery based on more than one data item. In contrast to most large-scale quantitative empirical applications of the capability approach that draw on available datasets (for example, the UN HDI described in Chapter One), this particular index enables information specifically on a community?s capabilities based on in-depth inquiry through primary data collection on indicators.  Indicators have been used to inform decision-making, improve stakeholder participation, build consensus, explore underlying processes and enhance advocacy (Parris and Kates, 2003)33.  One of the key resources in post-disaster work that has successfully used indicators is a handbook published by The Sphere Project (2004). The Sphere Project was initiated in 1997 to improve the quality of assistance provided after a disaster through a set of minimum standards.  The handbook describes indicators as ?signals? that show whether or not certain standards have been attained, and demonstrates how indicators can be effectively used in assessing outcomes and following trends (The Sphere Project, 2004).    Various indices to date demonstrate how multiple indicators can be combined to construct indices that aim to reduce the complexity of an entire system to a single metric (Tate, 2011).  Similarly, in this study, an index has been used due to its advantages as an efficient data-reduction device while allowing for several indicators to be summarized with a single numerical score, while also maintaining the details of all individual indicators (Babbie, 2004).  While recovery indices are not readily available, those of vulnerability are becoming more so.  Some examples include the Disaster Risk Index (UNDP, 2004), the Prevalent Vulnerability Index, and the Environmental Sustainability Index (Esty et al, 2005) and the Social Vulnerability Index (Cutter, 2008).  The design of a recovery measure ultimately depends on the desired features that are being measured and the data available to apply it (Rathfon, 2010).  The MDRI tool will ultimately allow for                                                    33 Development agencies actively use indicator tools in their work. For example, indicators are used for purposes of developing outputs and outcomes when designing projects and for monitoring and evaluation the progress of projects.  For example see risk reduction indicators developed in response to the Indian Ocean Tsunami in Provention, 2006.      148  the exploration of different relationships between specific dimensions of recovery and overall recovery, as well as resettlement characteristics in the next chapter. The index that has been constructed for the research follows guidance from existing methodological papers, and uses creative applications from social indicator research in planning, risk reduction and related disciplines.  However, in spite of its frequent use in social science research, the methodological literature contains little discussion on index construction (Ibid).  Most guidelines that are available rely on quantitative data, with few incorporating qualitative data.   However, there is wide acknowledgment that indexes can be successfully constructed from many different types of data and for a variety of purposes, and several indices that have systematically incorporated qualitative observational data (e.g., Brown et al., 2010; Cutter et al., 2010).   The detailed process of forming the composite index includes the following parts: selecting relevant capabilities, matching capabilities to appropriate indicators, rating indicators and index scoring. Each step will be described in detail and once the index is developed, it will be applied to the five village cases in order to determine capability-specific scores and a total recovery score for each case.  Based on the total recovery score, an overall holistic community recovery ranking across cases will be articulated.  This will be followed by an analysis of what the ranking findings indicate based on correlations across capabilities. 5.2. Selecting relevant capabilities  The first step in implementing the capabilities approach is to identify which capabilities will be taken into consideration. Capabilities were selected primarily on the basis of relevance, with each capability having a direct link to community recovery.  As Sen writes, ?The focus has to be on the underlying concerns and values, in terms of which some definable capabilities may be important and other quite trivial and negligible? (Sen, 1993: 32).   Ideally, as Sen (2005) and others argue, a central factor in the application of the capability approach is the open validation of a capabilities list, for example through public scrutiny and debate.  Most qualitative empirical techniques using the capabilities approach have utilized participatory methods for the selection of capabilities and/or functioning.  For example, in her work on three development projects in Pakistan, Alkire (2002) selects functionings by conducting community discussions on project impacts and conducts participant ranking of functionings and capability enhancing effects.  Due to time and scope limitations, this study did not directly use a method or  149  participatory process in which subjects were asked what their capabilities were and how relevant they were for them.  However, all of the capabilities chosen to compare villages were indirectly probed and articulated throughout the course of the fieldwork; this information was used to develop the final list. As such, the method used for developing a list of relevant capabilities for assessing recovery followed the three following consecutive steps of continuous modification:  Prior to embarking on fieldwork, a broad list of capabilities was prepared based on work in Central Human Capabilities by Nussbaum (2000), in which she defends a list of capabilities that are the moral entitlement of every human being.  This list was adapted based on preliminary observations from the exploratory trip.   While in the field, the list of capabilities was modified based on interactions and insights on the recovery process. Only those capabilities most relevant to the context of current day Aceh and the topic of disaster recovery were focused on (Sen, 1999).  The final list of capabilities and was completed during the data analysis. This list was a further modified version of the list used in step two, based on an analysis of contextual interview data and the accuracy of data.   The final list of capabilities that have been articulated for this study is as follows: 1. Life ? being able to live a long and complete life  2. Bodily health ? being able to have good health and adequate nourishment 3. Mental wellbeing ? being able to be mentally healthy 4. Education ? being able to be educated 5. Shelter ? being able to be sheltered 6. Mobility ? being able to be mobile and move freely 7. Social interaction ? being able to engage in social interaction 8. Attachments ? being able to have attachments to people and things 9. Environment ? being able to live with concern for natural surroundings 10. Livelihood ? being able to seek employment and attain a living 11. Exposure? being able to be protected from hazards and forces of nature 12. Respect ? being able to live free of discrimination, crime and violence Essentially, it is the twelve capabilities listed that form the items of the multidimensional composite recovery index.  The logical validity of each capability to the overall variable of recovery  150  was assumed since the theoretical framework has already demonstrated the relationship between capabilities and dimensions of recovery within a framework of development (Chapter Two).   Two capabilities that are recognized as significant in the recovery of Aceh, yet were excluded for purposes of measuring community recovery in this dissertation include the following:  Political ? being able to participate in political decisions Religion ? being able to observe religion freely  The political capability was excluded for several reasons because of the complexities associated with its measurement.  In the context of Aceh, it is hard to distinguish between post-tsunami recovery and post-conflict recovery.  Many have argued that one cannot be addressed without the other.  There are also sensitivities related to topics surrounding the Free Aceh Movement (described in Section 1.3.3) and the research aimed to steer clear of related discussions as some interviewees were previously closely associated with the movement as combatants. As such, data collection did not directly inquire into topics of political recovery status in significant depth.  The exclusion of the capability is warranted, as the unique post-conflict political environment is fairly consistent across all cases explored.  Gathering local voting data was explored, but not pursued due to challenges in data availability and accessibility. However, elements of involvement in local village politics is incorporated within the dimensions of social interaction.   The religion capability has also been excluded.  With the implementation of Shariah law, freedoms associated with religious practice are situated within a unique governance system.  In fact, it can be argued that many subjects value the practice of religion as the most important capability, while others have no choice but to value it as such.  Again, the exclusion of the capability is acceptable as the Shariah law and practice of Islam was consistent across all cases.  The essence of the capability is captured to some degree in the respect capability, with variables of tolerance and crime.   5.3. Matching selected capabilities to indicators  5.3.1. Community scale  At its core, the capability approach is an evaluation framework for individual welfare, and thus, in his work, Sen formally defines capabilities at an individual scale.  However, it can be argued that there is no restriction on the unit used for measuring capabilities as the general principles behind  151  the approach, including notions of freedom and choice, are undoubtedly relevant at a various scales. This idea has been exemplified in the Human Development Index, a capabilities based national measurement scheme, described in Chapter Two.  In addition, although the unit of analysis for capabilities is intrinsically the individual, applications at the community scale are feasible given it is a level that is clearly derived from and linked to individuals (Comim et al., 2008).  In the context of the developing world, including Aceh, where societies tend to be more collectivist than individualistic, an emphasis on capabilities manifested at the community scale is of particular importance and value.  At the same time, many capabilities are interdependent at higher scales, in which opportunities are open to all, but not everyone is able to realize them at the same time (Robeyns, 2006).  Exploring them at a community scale allows for incorporating some of these interdependencies and power relations within a given capability.     5.3.2. Indicator selection To operationalize the capabilities as measurable items, multiple indicators are developed at a community scale that can be used to represent each capability being measured. As the intention is to assess community-scale recovery, community indicators were developed. Indicators were all descriptive of the different elements of living conditions, within the background of recovery.  Since the capabilities are such that they do not have defined corresponding measures, a selection of indicators that could represent each capability were proposed and modified throughout the data collection process.  The general criteria used to ensure that indicators were appropriate and accurate measures of specific capabilities were: 1) the indicator was representative of the corresponding capability; and 2) it was intuitively plausible (Gardoni & Murphy, 2008).  In many cases, in order to move from individual capabilities to community indicators, key indicators for the capability were identified and then aggregated to all members of the community (i.e., the community scale).  Indicators were also chosen for their validity and reliability ? if it was a good indicator of what was being measured, and if it produced results that provided accurate information of what is being inquired. A table outlining the proposed relevance of each indicator used and recovery outcome is included in Appendix Two).  Other criteria used for selecting effective indicators included the availability and accessibility of information. For many initially proposed indicators, there was either no data available for the indicator and/or the data were inaccessible or unreliable.  Therefore, only those indicators that were feasible during data collection were included in the final list.  In many cases, proxy indicators were selected under the relevant capability. In addition,  152  attempts were made to include only those indicators that were fairly consistent across villages for pre-disaster conditions were included (so aspects such as income equality were omitted, for example, because some villages already had high inequalities before the disaster).  However, this was a difficult task, given the lack of information on the pre-disaster condition of many villages in Aceh.  Other criterion considered when possible included: the indicators potential for reproducibility (i.e., whether or not data could be independently verified and reproduced) and the robustness of the indicator in terms of how analytically sound it appeared in terms of its relationship to community recovery.  5.3.3. Indicator data sources The collection of primary data as a procedure for measuring capabilities has become increasingly frequent as applications of the approach develop (Comim et al., 2008: 177). The method used in this study follows this trend as indicator data is primarily based on primary data collection.  As described in the methodology chapter, data collection commenced using a tentative list of indicators derived from these capabilities and based on a review of literatures using indicators.  Concurrently, some preliminary indicators were eliminated and/or expanded upon based on data availability, relevance and feedback from local researchers (for example, university attendance rates and number of civic associations were omitted, and number of permanent plants was added). Most of the data was collected through semi-structured interviews with key informants and/or direct observations during fieldwork.  To ensure consistency and accuracy, all primary data that has been used in the assessment was collected during the official fieldwork period over five months in 2011.  Limited indicator data is derived from secondary sources, including village office documents (maps and records) and BPS data.  To ensure consistency in the time element of indicator data collection, the marker that has been used is data from at least 5 years from the disaster (i.e., 2010 onwards). Specific data that was used for each indicator is indicated in Table 5.21. The large amounts of data collected were systematically processed from the different data sources to the index. Two examples are provided. (1) Semi-structured interviews with the midwives in each village pointed to major categories of illnesses, namely: the prevalence or absence of serious illnesses (i.e. infectious disease and STDs), the prevalence of unique symptoms (i.e. asthma and malnutrition) and/or simply common systems (i.e. cold and fever) in the community.  The average number of patients a week was also recorded in order to assess illnesses relative to the population size.  Data from focus group discussion with villagers was used to support data on reported illness.   153  For example, in Panteriek Compound, it was common for community members to describe a higher occurrence of asthma, but no serious illness.  Two out of the five health centres had readily available documentation on number of visits per day, diagnosis and treatment.  While this data was collected, it was not used as a source of data for the index.  (2) Direct observational data that had been recorded over the course of time spent in a particular village described the daily life of the community and indicated the number of gatherings visible each day.  For example, this included descriptions of the types of people that had gathered, locations of the gatherings and the purpose of the gathering (if evident by observation).  These observations were processed as way of measuring the use of public space and as an indicator for social interaction.  Focus group discussions in each of the villages included discussion on chatting with community members, getting together on a regular basis, having space to hang out, and common meeting up spots. These data were used to support the direct observations recorded.    154    Capability IndicatorKey InformantFocus Group Discussions Direct ObservationsDocumentsliving conditions (in this case, last two years)x omarriage rates xbirth rates xpopulation change xreported illness x oavailable health services x xnumber of food vendors x xproductive cropland x xavailable clean water x xexpressions of grief x oexpressions of fear x onegative reminders of disaster (debris, destruction)xemphasis on continued education (high school +)x o oschool attendance levels o xsupplies/equipment availableo xquality of homes o xavailable land for new homeso xpopulation in semi-permanent/wood houses (as %)xquality of roads xpublic transit o opopulation with motorbikes (as %)o xnumber of community activities and/or eventsx ouse of public space o xcommunity satisfaction o xinvolvement in local village politicsxLife ? being able to live a long and complete lifeBodily health ? being able to have good health and adequate nourishmentMental wellbeing ? being able to be mentally healthyEducation ? being able to be educatedShelter ? being able to be shelteredMobility ? being able to be mobile and move freelySocial interaction ? being able to engage in social interactionx = central data source; o = subordinate data source 155    Table 5.1. Indicator data sources 5.4. Rating indicators and index scoring Appendix Three displays the complete MDRI.  The table provides a detailed representation of the community recovery index calculations and final scoring.  These are calculated through the following three steps of: 1) indicator scoring, 2) capability index scoring, and 3) recovery index scoring.  Capability IndicatorKey InformantFocus Group Discussions Direct ObservationsDocumentshome ownership xnumber of renovations xnumber of dong dong plantsxhome occupancy rate o xvegetation xlittering xgarbage disposal xunemployment rate o xtraditional survival options (fishponds, livestock)o xhome industries and small shopso xpreparedness planning and awareness x ohazards (flooding based on disaster event)o xdisaster risk (based on TDMRC)xlevel of tolerance (ethnic and religious)o xgender inequality o xcrime rate x odisparity between young adults and eldersx ox = central data source; o = subordinate data sourceAttachments ? being able to have attachments to people and thingsEnvironment ? being able to live with concern for Livelihood ? being able to seek employment and attain a livingExposure ? being able to be protected from hazards and forces of natureRespect ? being able to live free of discrimination, crime and violence  156  1) Indicator Score (X) As displayed in Table 5.2, many indicators are based on qualitative observational and interview data across a small sample size.  Consequently, indicator scoring had to accommodate a broad range of responses.  To enable measurement, each indicator was rated on a three-point scale (see Appendix Three).  In most cases, the scaling was developed through allocating three distinct parameters that covered the actual variation across cases; this was based on recovery data collected.  The three point scaling was chosen as it was the best way to cover the full spectrum of available scores across the distribution of cases, and ensure sufficient scores in each range.   This was done in order to overcome challenges in the lack of accurate records of detailed indicator data and variation in responses.   2) Capability Index Score (C) A capability index score is calculated by giving each indicator an equal weight within a given capability.34  This would allow for capabilities to be compared against each other in further analysis. In equation form:      [?      ]    Where  Cj = jth capability       (1)  N = number of indicators  Xi = ith indicator 3) Recovery Score (R) Before calculating the recovery score, it is important to determine the weighting of each capability.  For example, indicators that are considered more important for measuring the phenomenon of interest are given great weights (Booysen, 2002). This step of weighting index components has been described as being one of the most subjective decisions when constructing an index (Morse, 2004, Bohringer and Jochem, 2007, Tate, 2011).   On weighting capabilities, Sen writes, ?Along with the exercise of listing the relevant capabilities, there is also the problem of                                                    34 This follows an assumption that the villagers would do the same.  Relative weightings by villagers was not explored.  157  determining the relative weights and importance of the different capabilities included in the relevant list? (Sen, 2005: 158).  Initially, variable weighting options were considered in order to emphasis specific types of recovery (i.e., social recovery, long-term recovery, and so on).  However, the decision to use equal weighting was taken.  This was due to uncertainty in the appropriate importance of each capability due to an incomplete understanding of the underlying processes involved to assign meaningful weights (Cutter et al., 2003; Tate, 2011). In addition, equal weights would not necessarily bias the analysis and findings, especially as there was a lack of statistical or participatory approach feeding directly into the weighting process.  For these and other reasons, fixed weights are far more typical than variable weights when constructing indices (Tate, 2011).  Similar approaches have been justified by other researchers measuring capabilities in specific. For example:  The process of deliberation and judgment involved in evaluating capability sets does not need to be comprehensive or complete for practical purposes.  Weights do not need to be precisely specified to have practical significance; they can rather be confined to certain ranges or can cover merely the scope of the decision or of the judgment to be taken in a particular case.  (Comim et al., 2008: 184).  Aggregation refers to the method used to combine normalized and weighted indicators into the final index (Tate, 2011). Because each capability was weighted equally, aggregation was done by summation.  Therefore, the final recovery score is simply calculated by accumulating scores assigned to each individual capability, or in other words, a sum of all of the sub-indices.  As an equation:    ?       Where  R = recovery score        (2)  Cj = jth capability  M = number of capabilities It is important to note here that recovery (R) is a measure of relative well-being across cases, rather than a measure of each case post-disaster relative to its pre-disaster state.   158  5.5. Index validation The MDRI has been constructed with close attention in order to maximize validity (i.e., maximize the likelihood of the index actually measuring community recovery).  Techniques to do this include comparisons with other documented ways of measuring community recovery (i.e., Rubin 1985, Brown et al., 2010).  However, extensive validation of the index is difficult given the depth of the data required and inaccessibility of the case studies.  With the large number of villages in the region of study, there are few researchers (familiar to this researcher) in Aceh who know each of the village cases in enough depth to contribute to the validation.  In order to conduct a basic validation of the tool, the same research assistant (RA) who accompanied the researcher during fieldwork was asked to independently provide her village rankings on the different dimensions of recovery included in the index.  The rankings of the RA supported those of the author.  A final list of 42 indicators across the 12 capabilities is listed in Table 5.2.  The cases are ordered in increasing level of recovery (R).  In the table, X and C range from 1=low to 3=high, and R ranges from 12=lowest to 36=highest.  Mean scores are listed to enable comparisons in the next section.  Detailed breakdowns used to develop scores are included in Appendix Three. 159    Table 5.2. MDRI with scores X C X C X C X C X Cliving conditions (last two years) 2 1 2 3 3marriage rates 1 3 3 3 1birth rates 1 2 3 3 1population change 1 1.3 2 2.0 2 2.5 3 3.0 2 1.8reported illness 1 3 3 1 2available health services 2 2 1 3 1number of food vendors 1 1 2 3 3productive cropland 1 3 1 2 1available clean water 11.232.421.822.232.0expressions of grief 3 1 3 1 2expressions of fear 3 1 3 1 2negative reminders of disaster 3 3.0 2 1.3 3 3.0 1 1.0 3 2.3emphasis on continued education 2 1 1 2 3school attendance levels 1 2 2 2 3supplies/equipment available 11.321.721.732.322.7quality of homes 3 3 2 2 1available land for new homes 2 3 1 1 1population in semipermanent houses3 2.7 1 2.3 2 1.7 1 1.3 3 1.7quality of roads 3 3 2 1 2public transit 1 2 3 3 2population with motorbikes (as %) 1 1.7 2 2.3 1 2.0 3 2.3 3 2.3number of community activities 2 1 2 3 3use of public space 1 1 2 2 3community satisfaction 1 2 1 2 3involvement in local village politics 1 1.3 1 1.3 2 1.8 3 2.5 2 2.8home ownership 2 1 3 2 3number of renovations 1 2 1 2 3number of 'permanent' plants 1 3 1 2 3home occupancy rate 1 1.3 2 2.0 3 2.0 2 2.0 3 3.0vegetation 2 3 1 2 3litering 2 3 3 1 3garbage disposal 1 1.7 3 3.0 1 1.7 2 1.7 3 3.0unemployment rate 1 1 2 3 2traditional survival options 1 2 1 3 1home industries and small shops 1 1.0 1 1.3 2 1.7 3 3.0 2 1.7preparedness planning and awarness 1 3 1 2 1hazards (based on disaster event) 3 1 3 1 2disaster risk (based on TDMRC) 32.322.032.311.311.3level of tolerance (ethnic and religious)2 2 1 2 3gender inequality 2 1 1 2 3crime rate 2 2 3 2 1disparity between young adults and elders1 1.8 1 1.5 3 2.0 2 2.0 3 2.5Total Recovery Score (R)20.4 23.2 24.1 24.7 27.0Mean Score 1.7 1.9 2.0 2.1 2.3Highest capability in caseHighest case in capabilityGampong BaroLampuloPanteriek CompoundCapability IndicatorNeheun CompoundBitai1Life ? being able to live a long and complete life2Bodily health ? being able to have good health and adequate nourishment3Mental wellbeing ? being able to be mentally healthy4Education ? being able to be educated5Shelter ? being able to be sheltered6Mobility ? being able to be mobile and move freely7Social interaction ? being able to engage in social interaction8Attachments ? being able to have attachments to people and things9Environment ? being able to live with concern for natural surroundings10Livelihood ? being able to seek employment and attain a living11Exposure ? being able to be protected from hazards and forces of nature12Respect ? being able to live free of discrimination, crime and violence  160  5.6. Recovery ranking  Based on the MDRI, the overall holistic community recovery ranking of the five village cases from highest recovery to lowest recovery is: 1. Panteriek Compound (highest) 2. Lampulo  3. Gampong Baro 4. Bitai  5. Neheun Compound (lowest) The rankings listed above point to some compelling findings.  For instance, the two villages ranked highest overall in recovery differ in participation and movement.   If we refer back to our case selection table (Table 5.3 below), the rankings show that both the highest recovered case (Panteriek Compound) and the lowest recovered case (Neheun Compound) are 'moved' cases and that the two 'not moved' cases rank second and fourth in terms of overall recovery.  Therefore, rankings indicate that movement does not necessarily directly relate to overall recovery score.  The mixed ranking goes against the expectation that rootedness is directly related to positive recovery outcomes and findings point to the need to understand what other elements are at play.   In terms of participation, one non-participation case manifests the highest recovery score (Panteriek Compound) while one other has the lowest recovery score (Neheun Compound). The participatory cases rank second and third in recovery level.  Rankings indicate that a participatory process does not necessarily directly correlate to overall recovery score.  The mixed ranking of cases questions the assumption that higher participation is of utmost priority, but rather suggest the need to understand the different meanings of participation and balance participatory processes alongside other pressing needs.  The findings also highlight that it is not a matter of simply movement versus non-movement and/or participation versus non-participation, in terms or enabling positive recovery outcomes.    161    Table 5.3. Case criteria and communities selected (ranked) (#1 = highest recovery score) In addition, MDRI results point to some observations in relationship to basic dimensions of movement and participation that support the need to explore some of the other elements at play.   These findings will be further explored in the next chapter (Chapter Six).  For example:  ? Only two of the three resettlements in a new location, Neheun Compound and Gampong Baro, have a high score in mental wellbeing ? Lampulo and Panteriek Compound both ranked low on safety, even though one was a resettlement in