Regional Environmental Change https://doi.org/10.1007/s10113-017-1260-z

ORIGINAL ARTICLE

Long-term recovery narratives following major disasters in Southeast Asia Frank Thomalla 1 & Louis Lebel 2 & Michael Boyland 3 & Danny Marks 4,5 & Ham Kimkong 6 & Sinh Bach Tan 7 & Agus Nugroho 3 Received: 9 June 2017 / Accepted: 16 November 2017 # Springer-Verlag GmbH Germany, part of Springer Nature 2017

Abstract Most studies of major disasters focus on the impacts of the event and the short-term responses. Some evaluate the underlying causes of vulnerability, but few follow-up events years later to evaluate the consequences of early framings of the recovery process. The objective of this study was to improve understanding of the influence that recovery narratives have had on how decisions and actions are undertaken to recover from a disaster, and what influence this has had in turn, on long-term resilience. The study drew on comparisons and insights from four case studies in Southeast Asia: (1) local innovations that led to new policies for living with floods in the Mekong Delta in Vietnam following the 2001 Mekong River floods; (2) livelihood and infrastructure responses in Prey Veng, Cambodia, after the 2001 and 2011 Mekong River floods; (3) the role of the Panglima Laot, a traditional fisheries management institution, in the recovery process following the 2004 Indian Ocean tsunami in Aceh province, Indonesia; and (4) the challenges faced by small and medium enterprises in a market area following the 2011 floods in Bangkok, Thailand. This study identified alternative narratives on the purpose and means of ‘recovery’ with implications for who ultimately benefits and who remains at risk. The study also found both formal and informal loss and damage systems were involved in recoveries. The findings of this study are important for improving the performance of loss and damage systems, both existing and planned, and, ultimately, supporting more climate resilient development that is inclusive.

Keywords Disasters . Long-term recovery . Narratives . Loss and damage . Resilience . Southeast Asia

Editor:Robbert Biesbroek. Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10113-017-1260-z) contains supplementary material, which is available to authorized users. * Frank Thomalla [email protected]

1

Stockholm Environment Institute - Asia Centre, Chulalongkorn University, 15th Floor Witthayakit Building, 254 Chula Soi 64, Bangkok 10330, Thailand

Louis Lebel [email protected]

2

Unit for Social and Environmental Research, Chiang Mai University School of Public Policy, Chiang Mai 50200, Thailand

Michael Boyland [email protected]

3

Stockholm Environment Institute - Asia Centre, Bangkok, Thailand

4

Department of Asian and International Studies, City University of Hong Kong, Kowloon Tong, Hong Kong

Danny Marks [email protected]

5

Ham Kimkong [email protected]

Center for Social Development Studies, Chulalongkorn University, Bangkok, Thailand

6

Sinh Bach Tan [email protected]

Department of Natural Resource Management and Development, Royal University of Phnom Penh, Phnom Penh, Cambodia

7

Research Centre of Science and Technology Policy, National Institute for Science and Technology Policy and Strategy Studies (NISTPASS), Ministry of Science and Technology (MOST), Hanoi, Vietnam

Agus Nugroho [email protected]

F. Thomalla et al.

Introduction Major disasters trigger complex social, ecological and political responses that can be relatively short-lived and discrete, or continue to unravel and interact for a decade or more. The extent to which the recovery process is successful in building resilience of different social groups to future disasters seems to depend upon whether or not the post-disaster ‘windows of opportunity’ for more transformative change can be exploited, and if so by whom (Birkmann et al. 2010). Such opportunities are easier to grasp if you have prepared for them, have resources and are supported by a network or coalition (Olsson et al. 2006). One way to build coalitions and change beliefs is through attractive stories or narratives (Shanahan et al. 2011). Disaster recovery narratives may play an important role in building support for particular approaches to recovery. Four main disaster recovery narratives can be distilled from the literature. The early restoration (ER) narrative argues that it is important to replace lost assets and get lives back to normal as quickly as possible. It is a popular idea that fits the need for authority to be seen to be taking immediate action. In this narrative, disaster response and long-term recovery initiatives are kept separate (Ingram et al. 2006). The response to the 2010 earthquake in Haiti, for example, gave high priority to providing shelter quickly, creating new settlements, rather than having residents returning to a neighbourhood made safer and less at risk (Sanderson et al. 2014). The problem is not only for infrastructure, however. After the 2006 earthquake in Yogyakarta, livelihoods which were insecure predisaster were re-habilitated rather than replaced (Joakim and Wismer 2015). The problem with this short-term approach is that it can easily re-create the social and economic vulnerabilities which contributed to the disaster in the first place. The linking relief, rehabilitation and development (LRRD) narrative focuses upon reducing the gaps between humanitarian aid and development cooperation (Mosel and Levine 2014). Doing so should reduce the trade-offs between addressing short-term needs and longer-term development aspirations (IOB 2013). This narrative is articulated primarily by aid agencies, the heroes, while blame is laid upon the hazard. While the goal sounds attractive, it has been difficult to put into practice (Audet 2015). One reason is that development assistance does not always follow humanitarian aid; another is that the various activities may occur in parallel rather than as a series of steps. The build back better (BBB) narrative—the guiding principle of the 2004 Indian Ocean tsunami recovery—as put by Bill Clinton, UN Special Envoy for Tsunami Recovery, promises to make sure that the recovery process accomplishes more than just restoring what was there before (Fan 2013). It can mean placing more emphasis on environmental sustainability, improved livelihood choices, ending civil conflicts and ensuring a fairer distribution of both risk and assistance, but it also

might mean stronger buildings and higher floodwalls or other technical fixes. The attraction of this narrative is that it can mean almost anything, as different actors define ‘better’ in their own ways to fit their own agendas (ibid). This narrative is used by a range of stakeholders, and thus, there are many heroes. One constraint is that an individual may feel more secure or happy with a return to what they were doing before, i.e. as the ER narrative promises (Daly 2015); another is that the power structures which make some groups vulnerable may not be easy to alter and thus constraint efforts at making things ‘better’, as shown in case of northern and eastern Sri Lanka after the 2004 tsunami (Khasalamwa 2009). In the last decade, a number of scholars (e.g. Klein 2007; Gotham 2008; Gunewardena and Schuller 2008) have argued that governments have used the BBB narrative to promote private neoliberal, capitalist interests, which affected populations might be less likely to accept under normal conditions. Thus, disaster recoveries have become opportunities for profit-making and corporate restructuring of societies. Klein (2007) used the term ‘disaster capitalism’ to encapsulate these post-disaster practices. Loewenstein (2015) details numerous examples of post-disaster profiteering by international corporations, most notably after the 2010 earthquake in Haiti. The empower local communities (ELC) narrative treats disaster recovery as a political process in which local communities need to be empowered so that their underlying vulnerabilities and root causes of disasters will be addressed. The state and corporations are typically portrayed as villains, and nongovernmental organisations (NGOs) and communities as heroes. Blame is laid on socio-economic structures. Close attention is paid to the interaction between aid agencies and local communities (Daly 2015). One problem is that altering power structures can be quite challenging; another is that actions at higher levels and with the state may also be needed. The recovery of New Orleans after Hurricane Katrina, for instance, was hampered by the lack of elites committed to the city that could have provided the necessary leadership for an effective recovery (Hobor 2015). It is not clear what influence these common narratives have on recovery policy and practices, or longer-term outcomes in specific contexts. Most studies of major disasters focus on the impacts of the event and the short-term responses. Some evaluate the root causes of vulnerability, but few conduct longitudinal studies or follow-up events years later to evaluate the consequences of ad hoc and more systematic, technical, political and social responses to the event. Various examples suggest that there is high value for the investigation of long-term disaster recovery processes. The objective of this study is to improve understanding of the influence that recovery narratives have had on how decisions and actions are undertaken to recover from a disaster, and what influence this has had, in turn, on long-term resilience building. By resilience we mean ‘the ability of a system,

Long-term recovery narratives following major disasters in Southeast Asia

community or society exposed to hazards to resist, absorb, accommodate, adapt to, transform and recover from the effects of a hazard in a timely and efficient manner, including through the preservation and restoration of its essential basic structures and functions through risk management.’ (UNGA 2016, p. 22). This is a broad, encompassing definition with many terms open to multiple interpretations and, thus, relevant to all four narratives. The use and influence of disaster recovery narratives are anticipated to vary with the type or impacts of the disaster, the availability of resources, institutional capacity and governance context. Major disasters may result in complex and protracted recovery phases that allow time for narratives to evolve and perhaps even for counter narratives to arise. With a lot of resources leveraged, the expectations and incentives to deploy ambitious narratives like BBB are higher than when resources for recovery are scarce. Political imperatives tend to favour short-term narratives like ER, whereas external agents may have the luxury of arguing a longer-term, LRRD approach. Strong capacities to govern favour narratives which include the state as a responsible actor, whereas weak capacities favour narratives centred on self-organisation at the community level, like ELC. To help explore these interactions, this paper builds on insights from four empirical case studies of recovery from major disasters in Southeast Asia, and published literature. In order to analyse recovery narratives, we evaluate a range of specific recovery interventions deployed following major disasters—both those under the umbrella of each narrative and those outside the definitions and principles of the narratives. This we refer to as a disaster loss and damage system. To focus the analysis of the consequences of recovery narratives for resilience in particular places, special attention is given to the loss and damage systems invoked or ignored under these narratives. Here, a loss and damage system is defined as the formal and informal institutions, processes and systematic actions which aim to assist communities and societies absorb, cope with, adapt to and recover from adverse effects of disasters that may be either irreversible (loss) or replaceable (damage). We use this term because we believe a better

understanding of how existing loss and damage systems perform is necessary for dealing with long-term impacts of climate change (James et al. 2014; Surminski and Lopez 2014), which may well increase the occurrence of major disasters in the region (Cardona et al. 2012).

Methods Research was conducted between 2014 and 2016 with qualitative research approach consisting of a literature review and document analysis, interviews with key recovery actors in each case study and four workshops. The review of academic literature and documents published by government, NGOs, civil society organisations (CSOs) and the mass media was done to establish the status of research on approaches to disaster recovery, such as LRRD, BBB, ER, ELC, long-term recovery and resilience, and to identify, collect and review relevant documents for each of the four case studies in order to develop the context for the research and to understand what is already known about the case studies and wider topic of disaster recovery. The literature review was conducted using the internet search engines Google and Google Scholar and the scientific database ScienceDirect. Collected digital copies of documents relevant to the project were archived in Zotero for author access. Both documents were written in English, and documents written in the local languages of the case study sites were included in the search. Case studies were selected for in-depth empirical research on recovery processes following four major disasters in the region occurring this century (Table 1; disaster data from CRED and Guha-Sapir 2017). The case of the 2000 flood in An Giang Province in the Mekong Delta of Vietnam focused on new approaches to dealing with flood risks. The 2001 flood in Cambodia case study focused on the resilience-building efforts in affected communities in Prey Veng Province that had also been impacted by major floods the previous year. The 2004 Indian Ocean tsunami case examined the role of the Panglima Laot (a customary fisheries institution) in the long-term recovery of fisheries communities in Aceh,

Table 1 Overview of disaster recovery case studies, including key loss and damage data (at the national level) and recovery narratives (disaster data from CRED and Guha-Sapir 2017). ELC empower local communities, BBB build back better, ER early restoration, LRRD linking relief, rehabilitation and development Case study title

Country

Disaster type

Year

Deaths

Total affected

Economic losses and damage Recovery (US$) narratives

Living with flood strategies in An Giang Local recovery programmes in Prey Veng The role of Panglima Laot in Aceh’s recovery SMEs’ recovery in Nonthaburi

Vietnam

Flood

2000

496

5 million

US$ 265 million

ELC, BBB

Cambodia Flood

2000–2001 403

5 million

US$ 175 million

ER

Indonesia Tsunami

2004

165,800 700,000

US$4.5 billion

LRRD, BBB

Thailand

2011

877

US$40 billion

ER, BBB

Flood

10 million

F. Thomalla et al.

Indonesia. The 2011 floods in Central Thailand case focused on the experience of small and medium enterprises (SMEs) in Bang Bua Thong (BBT) market in Nonthaburi Province. The four disaster case studies represent a range of different characteristics in terms of hazard types (floods and tsunami), scale of impacts, governance systems, social and economic development status, risk, vulnerability, disaster loss and damage systems and recovery narratives. A provisional case study protocol was jointly developed by team members as part of the research design process and revised again when comparisons began, to guide each iteration of data collection and analysis. The protocol defined key terms to ensure consistency of use, included guidelines on how to conduct literature reviews and empirical data collection, and identified research questions to be answered in line with the overall study objective. The protocol was then translated into the local languages used in the case study sites. This ensured a common overarching methodology across all case studies, while allowing flexibility to explore aspects important and specific to a case study. The questions first covered the features of the loss and damage systems involved in each recovery, as well as stakeholder expectations of what the recovery should achieve, and evidence of its effectiveness and efficiency. This was followed by a consideration of the role of disaster narratives in framing recovery problems, solutions and evaluations. At least 15 key informant interviews were conducted in each case study. Informants included researchers and experts that have written or talked about the recovery process or are otherwise known to have followed it closely; decision-makers and planners that had important roles in the recovery process with government authorities; influential individuals in donor, humanitarian aid, international NGOs, development banks driving or supporting key interventions in the recovery process; local community leaders or members of domestic CSOs instrumental in the recovery process as lobbyists, development partners, or selforganised community-based action takers, local business, chambers of commerce, business associations, livelihood associations, insurance firms, construction contractors, involved in recovery process; and affected local households. Ultimately, through comparisons and synthesis, the aim of the study was to better understand the aspects of loss and damage systems associated with both successes and failures in building resilience of the most affected through recovery processes.

Framing of problems and solutions Disaster recovery narratives are constructed by coalitions of stakeholders to help articulate particular combinations of problems and solutions as logical courses of action that should be followed. Thus, arguments about the lack of protective infrastructure being the cause of flood risk, vulnerability and impacts lead to solutions involving the construction of physical infrastructure with governments and/or international donors stepping in as the heroes. Where the problem is identified as destroyed livelihoods and local institutions, as in the case of Aceh, the solutions include providing skills and resources to rebuild those livelihoods and building capacity of local institutions, typically with a role for CSOs. One limitation of the BBB narrative in Aceh is that it requires organisations with experience primarily in emergency relief to tackle recovery and development problems, areas in which they may not have sufficient expertise. Recovery policies primarily driven by the ER narrative, as in Prey Veng, may not address longer-term needs nor contribute to the building of resilience.

Proponents of narratives The recovery narratives of donors and national governments often significantly differed than those of local authorities and communities on the ground. In Aceh, the BBB narrative was prominent in the international arena, whereas with local governments and communities, the primary concern was with returning to normal as quickly as possible, as per the ER narrative. After the 2011 floods in Thailand, authorities used a combination of the ER and BBB narratives to justify public investments in protective infrastructure for key commercial areas and industry. On the other hand, several SME and BBT residents did not consider a serious flood as being a serious future risk or felt better prepared to cope with future floods, so their concerns were more focused on local economic development and political accountability—in line with the ELC and LRRD narratives. In the case of the residential cluster programme in the An Giang case, the views of authorities and residents about the residential cluster approach converged over time but were initially quite different. Residents that have had to move into the clusters were dissatisfied with the level of services provided, and it was only after a few years that support for the local ‘living with flood’ variant of the BBB approach pursued by the government gained wide acceptance.

Influence of recovery aims and outcomes

Disaster recovery narratives In this section, we present examples of how the four narratives described in the introduction (ELC, BBB, ER and LRRD) and their characters, proponents and influence have been used in the recovery process in each of the four case studies.

Recovery narratives can also evolve over time, underlining that recovery is not a distinct and discreet disaster management phase, nor a strictly linear process. For instance, in Aceh, one of the narratives of the response (i.e. roughly first 0– 18 months) was LRRD (Christoplos 2006), which was

Long-term recovery narratives following major disasters in Southeast Asia

replaced by BBB despite their many overlaps. However, overlapping narratives may actually guide the difficult transition from response to recovery, if they are complimentary and not conflicting. ‘Build back better’ was intended to emphasise the need for mainstreaming risk reduction into recovery efforts, but instead became a ‘catch-all’ phrase for humanitarian and development actors alike, despite their differing primary goals. The changing of post-disaster narratives will not necessarily align with changes in actions on the ground (i.e. from response to recovery/resilience-building), demonstrating the mismatch between international actors’ and local communities’ narratives. A certain degree of ambiguity is important for building coalitions around a recovery narrative. In the case of BBB, the vague definitions and principles, and multiple different uses of the term by various actors to refer to response, recovery and development initiatives mean that it is not evident what ‘better’ looks like, who it is better for and how exactly is it better than what existed before. In Thailand’s flood recovery, big multi-national businesses were given target assistance by the national government and thus were better off; meanwhile, local SMEs were excluded from many recovery efforts. More broadly, those who took precautionary measures (e.g. flood insurance) and were better prepared for the floods had the means to recover more efficiently and effectively. These factors need to be accounted for in recovery planning so that those who lack the capacity to recover, or become ‘better off’ as a result, can be targeted by interventions. In Nonthaburi, there was a mismatch between those who were worst affected and had the fewest assets to recover on their own, and those who were the target beneficiaries of government recovery assistance. This high degree of ambiguity and inequality allows governments to continue pre-disaster patterns of highly uneven socio-economic development. However, if differing stakeholder goals can be captured within one agreeable narrative, it can play an important role in arranging coalitions. The dominant narrative in three cases was driven by elite actors to build coalitions, sometimes serving their own narrow interests and sometimes adopting a paternalistic approach to meeting the needs of vulnerable groups and affected communities. Thus, in the Bangkok case, the ER narrative was initially used to try and garner support for raising roads and supporting the early recovery of the automobile industry. The government itself had to show it was doing something to protect foreign investments as well as regain lost political support from residents. In practice, residents living outside the core protected areas of Bangkok and industrial parks remained highly sceptical of the proposed measures even when supported with a BBB narrative, so they remained wary. However, SME owners did not consider it likely that such a serious flood would happen again, and so they were not particularly concerned with ‘building back better’, or indeed the role of changing recovery narratives. Perceptions of a

future threat have an important bearing on what affected people will want from recovery, yet it is more than simply being affected that influences risk perception. In An Giang, a living with flood discourse consistent with an ecologically oriented BBB narrative turned out to have many positive long-term consequences for land and water management that local authorities and residents did not initially accept or even anticipate. In Prey Veng, the ER narrative dominated and was reflected in quick-fix approaches to the construction of bridges and roads. Such a narrative encouraged quick action and the appearance of having done something rather than investing in high-quality infrastructure and land-use planning that could greater reduce the risks of floods for years to come.

Victims, heroes and villains—characters in recovery narratives In Aceh, there was a much stronger presence of international donors and actors. The BBB narrative was prominent at the national level and among these international actors. In a way, the narrative became a boundary object around which those with funds and those needing funds to rebuild, could negotiate and agree on reconstruction priorities. In this case, narratives portray donors, aid agencies and the international community (e.g. UN agencies) as heroes and local communities as victims. The national government’s coordinating Agency of the Rehabilitation and Reconstruction (BRR) was at times portrayed as both a hero and a villain, including by NGOs critical of the state’s slow reaction to the disaster. There were also the criticisms of NGOs from marginalised communities for only dealing with village officials whose leadership was not widely trusted, leading to unintentionally exacerbating inequalities and tensions (Telford and Cosgrave 2006). The Acehnese civil conflict, which had kept the province closed to outsiders for decades, created pre-existing vulnerabilities, compounded the impact of the disaster and at times led to inequalities in the recovery process itself (Telford and Cosgrave 2006). The international community were heralded for enabling an end to the conflict, but lambasted for not addressing the long-term social and economic impacts of the conflict through recovery initiatives—many chose to address only disaster impacts and not conflict impacts.

Recovery by whom and for whom? The dominant narratives of the four case studies offer indicative examples of how disaster recovery phases can differ, in part, due to who applies the discourse. For instance, the BBB recovery narrative used in Aceh encapsulates most, if not all, guiding principles of an efficient and effective disaster recovery, to the extent that at the same time the term can mean anything to anyone, and also offer no common concise definition across examples of its use (Fan 2013). The first explicit

F. Thomalla et al.

articulation of ‘build back better’ and its 10 key propositions came in the wake of the Indian Ocean tsunami by Bill Clinton (UN 2006) and offers important approaches to sound programming, across the response-reconstruction-rehabilitationrecovery-resilience-development time spectrum, which are not actually specific to build back better, nor critical of its assumptions. It has since been used as the dominate narrative in Myanmar following Cyclone Nargis in 2008 and in Haiti following the earthquake in 2010 (Fan 2013). The build back better narrative was conceived, promoted and is largely still used most by United Nations agencies (i.e. UNISDR) and policies (i.e. Hyogo Framework for Action 2005-2015 (HFA) and Sendai Framework for Disaster Risk Reduction 2015-2030 (Sendai Framework)). Limited prior analysis has been carried out on how various humanitarian, recovery and development actors across multiple scales (donors, INGOs, NGOs, national to local governments, CSOs and community-based institutions) translate and implement the UN principles for programmes and into on-theground action. In the case of Aceh, the overarching recovery goal of local actors such as the customary fisheries communities’ institution of the Panglima Laot was actually to return local conditions—such as infrastructure, housing and livelihoods—back to ‘normal’ (i.e. pre-tsunami status) as quickly as possible, rather than aspiring to build a ‘better’ community or Aceh. Furthermore, the funding for building back better was separated out and estimated as an add-on to rebuilding the assets directed lost and damaged as a result of the disaster (World Bank 2006). Build back better was not mainstreamed into reconstruction efforts, but rather presented as an optional extra set of interventions that could be pursued because the recovery had available, more financial resources than were required for simply rebuilding the damage. Narratives, along with control of resources and power, help determine who sets the goals for recovery. For BBB, there are concerns around ownership of the goals, relevance for local actors and recovering communities, and substance—the UN propositions do not contain specific directives and so the question of who decides what to prioritise and how to translate ‘build back better’ into meaningful interventions is left open. At the local/implementation level in Aceh, the Panglima Laot were seeking to harmonise two overarching recovery goals: build back better (promoted by the UN, international actors and the BRR) and returning back to normal as quickly as possible (largely wished for by surviving local communities). The BRR were tasked with coordinating different humanitarian and development actors and initiatives, as well as considering the voices of the survivors through local organisations, under the umbrella principles of build back better and the vision of ‘Building (a) reliable, dignified, prosperous and democratic Aceh’. The BRR existed as a ministerial-level agency under President Yudhoyono’s direct authority. The BRR institutional model has been pursued and learnt from in

other disaster recoveries, namely Cyclone Nargis in Myanmar and the earthquake in Haiti. In Thailand, a centralised, autocratic approach to recovery decision-making was also taken; this is in conflict with both local governments who wanted more autonomy and affected people who increasingly took measures into their own hands in response to a growing lack of trust in the national government. BBB has emerged as the dominant recovery narrative at the global level, driven by the HFA and the Sendai Framework. Given the recent adoption of the Sendai Framework, and the move towards implementing the framework in vulnerable regions and nations, a more critical eye is needed for the call to ‘build back better in recovery, rehabilitation and reconstruction’ (UN General Assembly 2015, p. 21) (Table 2).

Loss and damage systems involved in post-disaster recovery Disaster recovery narratives provide a frame for preferentially promoting particular loss and damage systems and approaches. The loss and damage systems involved in four recovery case studies have been classified into five types: livelihoods, infrastructure, financial, institutional and cultural, and ecological (Table 3). In the following sections, we expand on the design features of some of the more important examples, paying special attention to whom they target, when they were active and how they were governed. We also consider the consistency with dominant and alternative disaster recovery narratives.

Livelihood or sector Some type of livelihood or sector stimulus policy was pursued by all of the national governments of the case study countries. For example, in Thailand, efforts focused in particular on supporting the automobile manufacturing industry, and lower income groups buying their first car through the new car tax rebate. Reducing interest rates, raising the minimum wage and the salaries of government officials, tax exemptions to foreign investors and the construction of flood protection infrastructure implemented through the flood management plan were other measures taken by the Government of Thailand to enable economic recovery. Overall, whenever the ER narrative dominated, programme activities emphasised early restoration of the original livelihoods or support for affected sectors, whereas those in BBB or LRRD narratives tried to introduce more alternatives. In practice, there were some cross-overs in approach among case studies, and a significant fraction of the efforts to introduce alternatives failed as local residents chose to return to their original livelihood activities.

Battle between municipal and national government Outer residents remain wary

Often ignored by local communities that emphasised return to normal

Bangkok Metropolitan Administration

United Nations, donors, national government (i.e. BRR)

Victim: uter Bangkok Villain: national government Hero: Municipal government Victim: fisherfolk Hero: Panglima Laot, NGOs Villain: NGOs

ELC, BBB

ER, BBB

LRRD, BBB

An Giang, Vietnam

Nonthaburi, Thailand

Aceh, Indonesia

Fisheries sector needs to be revitalised with quick replacement of lost boats and longer-term diversification of livelihood activities in value chains

National and provincial government Accepted by local communities after initial resistance Victim: residents Hero: government

ELC, BBB An Giang, Vietnam

Resident clusters allow protection of homes and critical buildings while still allowing room for water when it floods Inner Bangkok must be protected at all costs because of its national importance, so must strengthen protective infrastructure and water management rules

National and provincial government Accepted by localcommunities but with contestation Floods bring benefits too, so it is important to Victim: farmers allow dykes to overflow at least once every 3 years Hero: government

Influence Proponent Characters Recovery narratives

Example of specific narrative instance

Infrastructure

Case

Table 2 Examples of the use of narratives and their characters, proponents and influence in the case studies. ELC empower local communities, BBB build back better, ER early restoration, LRRD linking relief, rehabilitation and development

Long-term recovery narratives following major disasters in Southeast Asia

Losses and damages in infrastructure can be addressed through reconstruction funds, protective infrastructure projects, community-based institutions and household building improvements. In Indonesia, for example, a dedicated reconstruction fund was established by the BRR, UN, INGOs and a number of donors, with the aim to build back better. All case studies show that at the household level, people have taken measures to reduce losses and damages by undertaking improvements to their homes and businesses. In Indonesia, property ownership was improved by giving home owners whose houses were destroyed official titles, which most didn’t have before. However, challenges arose in establishing who owned which land after the houses and portions of land were washed away. The extent to which the BBB or LRRD narratives could be pursued in practice depended on the institutional arrangements made for dealing with loss and damages to infrastructure, including budget allocations for re-construction.

Financial Financial loss and damage systems included compensation, insurance and community household funds and loans. Compensation mechanisms existed in all countries. In Vietnam, compensation for household affected by disaster were available at the local level for a death or injury in the family, and for damages to or losses of property and crops. Banks provided loans to farmers for land purchases and house construction. Financial systems for handling loss and damages, based at government, donor, community or household level, largely reflected the logic of the ER narrative even when the dominant recovery narrative in a particular case study was BBB or LRRD.

Institutional and cultural Non-material loss and damage systems include culturalsocial support (institutions) and peace building and conflict resolution. For example, in Cambodia, as in other countries, there is a culture of solidarity and reciprocity that helped people in need. The social and cultural dimensions of loss and damage systems that emerged as significant in the different case studies were not always closely aligned with the logic of the dominant disaster recovery narrative. In particular, many of the locally oriented initiatives were more consistent with the minority ELC narrative than what was articulated by government or mainstream donors.

F. Thomalla et al. Table 3 Illustrative examples of different types of loss and damage systems involved in post-disaster recovery identified in the case studies. VN Vietnam, IN Indonesia, TH Thailand, CA Cambodia Loss and damage system type

Examples

Type 1: livelihood or sector Rehabilitation

Stimulus Transformation

Additional income from off-farm activities and growing additional crops (VN); diversification of economic activities along the fisheries P2C chain (IN) Replacement of fishing boats and gear within a year (IN) Tax exemptions to foreign investors and car tax rebate stimulus for auto-manufacturers (TH); reduction of interest rates, raising of minimum wage and salaries of government officials (TH) Migration to cities (CA) Trainings on alternatives to 3rd rice crop to better live with high flood risks (VN); alternative livelihood training (CA); trainings in business skills for SMEs (IN); changes in property ownership (IN)

Type 2: infrastructure Reconstruction Protection

Household rebuilding Type 3: financial Compensation Insurance Community funds Loans

Upgrading, relocation, raising of roads (IN, TH, CA) Flood protection infrastructure, raised roads, walls around industrial estates, flood embankments along waterways (TH); full and semi-dyke system (VN) Tsunami early warning system, evacuation routes and centres (IN) Structural improvements to houses and businesses (all); changes to construction techniques, siting and elevation (CA) Household compensation for death, injury, property and crop loss (CA) Dedicated reconstruction fund (IN) Flood insurance for SMEs (TH) Cash transfer, cash for work program, revolving fund for fisheries communities (IN); food for work program (CA) Borrowing from money lenders and micro finance schemes (CA) Bank loans to farmers for land purchases and house construction (VN)

Type 4: institutional and cultural Professional associations Community-based institutions/customs

Conflict resolution

Market association (TH) Panglima Laot customary institution (IN); disaster association (CA) gotong royong (mutual cooperation) (IN); culture of solidarity and reciprocity (CA); Phuong cham 4 tai cho (VN) Cultural memory and local knowledge (VN) Residential cluster programme (CA) Community disaster preparedness plans (TH) Community disaster preparedness training (IN) Aceh peace agreement (IN) Military coup (TH)

Type 5: ecological Ecosystem services for disaster risk reduction (DRR) Ecosystem services for livelihoods

Coastal exclusion zone (IN) Floodwater retention areas (TH) Accommodation, rather than control of floods (VN) Improved soil management (VN) Protection of coastal mangroves and marine ecosystems (IN) Diversification of cropping varieties (CA)

Ecological Ecological loss and damage systems pertain to the provision of ecosystem services for disaster risk reduction livelihood support and were relevant in all case studies. In Thailand, one issue was that few natural floodwater retention areas, such as wetlands and floodways, existed at the time of the 2011

flood (Marks 2015). More are being built now, but for the purpose of reducing water shortages during droughts. Although some elements of ecosystem-based approaches to recovery were found in most cases, it is only under BBB recovery narratives where there was some degree of fit between the solution framing of the narrative and significant and systematic loss and damage actions.

Long-term recovery narratives following major disasters in Southeast Asia

Overall, this section suggests that the multiple types of loss and damage systems, which overlap with each other, need to be considered when analysing recovery.

Improving the performance of loss and damage systems Our findings from the case studies show that the loss and damage systems that emerged in the years after each disaster had numerous limitations. First, there was limited downward accountability in these systems. For example, in the Thai case, the BBT mayor performed poorly during and after the floods, and has made no efforts to reduce the risk of future floods in the market area. Despite this failure, the mayor still remains in his position today because elections were suspended after the 2014 coup (Marks and Thomalla 2017). One reason for the limited level of accountability in these countries is that their governance structures remain centralised. National governments have retained too much power and resources in order to protect their interests. However, some of these national government agencies have performed poorly. For example, in Cambodia, they have only completed sections of roads, which might help to reduce the risk of some but not all future floods. Another problem is the fragmented governance of these L&D systems: government agencies still mostly work in silos and rarely collaborate with each other. One reason is because they want to protect their budgets (Marks and Lebel 2016). Further, the capacity of agencies working on DRR remains weak, with national leaders focused more on disaster response rather than prevention, despite these disasters heavily affecting their country. As a result, limited financial and human resources have been devoted to DRR. For example, in Vietnam, while agencies have provided houses for those who need to relocate due to the risk of high floods, they have been slow to provide essential services such as water and electricity to these houses. As a result of these governance shortcomings in the DRR sector, early warning systems in these countries are still limited and do not function well (Thomalla and Larsen 2010). For example, in Thailand, people received flood warnings too late—and presently, improvements have been limited. Another key barrier is that government agencies have been excluding key groups as beneficiaries in these systems. One group is the poor, whereby there are few compensation mechanisms and other safety nets designated for them if disasters strike again. For example, in Vietnam, smallholder farmers in the protected areas are not given enough support to help them live with floods, which requires water to be allowed to inundate the protected areas once every 3 years. Another excluded group is SMEs, despite the major role they play in each of these countries’ economies. For example, in Thailand, after the floods, the national government designed a set of policies

to help large businesses recover but did not target SMEs in their stimulus package. Given these limitations in these systems, a question that begs to be asked is how can these systems be improved. Given our findings from these cases, a number of crosscutting suggestions emerge. First, legal foundations of DRR policies and practices need to be more institutionalised since they are currently weak in these countries. Second, a centralised system is not necessarily the best type to reduce L&D because—as what happened during these disasters— such a system can make mistakes and unfairly distribute risks. Instead, a more devolved, participatory and cross-level decision-making system is needed. Communities whose risk to disasters would be increased or livelihoods adversely affected by these projects, such as floodwalls and new roads, should particularly be included in the decision-making process. Third, inter-agency and scalar fragmentation needs to be reduced and land use and water planning must be done collaboratively. Smaller DRR agencies particularly those in different ministries could be merged together. Financial and promotion incentives could be changed in order to encourage collaboration. Fourth, agencies involved in recovery and reducing the risk of future disasters need to keep in mind that BBB and other recovery narratives mean not only structural measures (more dykes¸ roads and dams) which sometimes only transfer future disaster risk rather than reduce it, but also non-structural ones as well (land-use planning, training programs, safety nets and better early warning systems). Fifth, at the same time, given DRR agencies’ limitations, heavily affected groups (SMEs, the urban poor and smallholder farmers) should not rely entirely upon them to improve L&D systems. Instead, they should self-organise to create their own early warning systems, provide social safety nets to those worst affected, become involved in planning processes and advocate for targeted assistance during disasters.

Discussion The case studies reveal the presence of different discourses on the purpose of ‘recovery’, ‘building back better’ and ‘resilience’. This has implications for who ultimately benefits from post-disaster recovery efforts and who remains at risk. Because the dominant narratives around recovery tend to be driven by elite actors that control resources and power, they set the goals for recovery, maintaining or strengthening existing power structures or creating new ones that serve their interests. This is particularly evident in Thailand, where the decision by the government to protect foreign-owned suburban manufacturing businesses and the urban elite in Bangkok’s city centre through the construction of flood protection walls increased the risk of people and small businesses located outside the areas protected by the walls. Here, the aim

F. Thomalla et al.

of recovery of the government was to appease big business and the urban elite in order to secure foreign investment, maintain economic activity and gain political favour from wealthy residents. The implications of these flood protection measures on the spatial and temporal distribution of risk across all Bangkok residents and businesses were not fully considered (Marks 2015). This is also evident in the Indonesia case, whereby the BBB narrative—constructed and championed from the top by UN agencies and other international actors—was not fully compatible with the tsunami recovery goals of those affected and literally recovering (Boyland et al. 2017). This example from Thailand also shows that the objectives of post-disaster actions (recovery, building back better, resilience) are not the same for all stakeholders. The objectives of SMEs affected by the 2011 Bangkok flood—and now arguably exposed to higher risk by the flood protection infrastructure built by the government since the event—were to restore their damaged businesses, to replace damaged merchandise and to protect and revitalise the local market area. Such a situation of different objectives of different stakeholders can influence the identification of villain and hero characters in narratives referring to contestation, such as local versus state, or urban elites versus the sub-urban poor. Disaster recovery and resilience building efforts need to consider who is likely to benefit and who is likely to lose from the outcomes of resilience building measures. Important questions to ask in the design of policies, strategies and activities are as follows: Who is the target of the planned intervention? Why are these particular individuals, social groups, businesses, etc. targeted? How will their resilience be strengthened? What potential negative affects might this have on other stakeholders now or in the future? Because the BBB narrative means different things to different people, it is also important to ask what is meant by ‘better’ and who defines this and why (Fan 2013). For all post-disaster narratives, we need to critically question the values, interests and assumptions that underpin planned initiatives or lack of action. These critical questions can be used to enhance monitoring and evaluation methods for determining the longer-term impacts of resilience building efforts. The findings are also relevant to current efforts to increase understanding of the links between disaster risk reduction, climate change adaptation, and loss and damage in theory and practice. The establishment of the Warsaw International Mechanism in 2013 (UNFCCC 2013) and the creation of a separate article in the Paris Agreement (UNFCCC 2015) have placed loss and damage firmly in the United Nations Framework Convention on Climate Change (UNFCCC). However, there is a broad spectrum of perspectives on loss and damage, and further research is needed to improve understanding of loss and damage systems in order to inform adaptation policy and decision-making (Boyd et al. 2017). This research contributes to this by

underlining the important role played by narratives in simplifying complex realities and building coalitions of support for particular approaches or courses of action. Another consideration is the extent to which resilience building efforts are truly transformative, even when they have been branded as such by disaster recovery narratives like BBB or LRRD. While the post-disaster recovery process has been described as a window of opportunity for transformative changes to be made, the evidence from our case studies indicate that none of the approaches employed in these locations can be described as transformative. Arguably, the most radical change of all the case studies occurred in An Giang Province in Vietnam, where the flood risk management policy changed from flood control to ‘living with floods’. However, the effectiveness of the approach is limited in the face of increasing hazard risks in the Mekong Delta and the narrative has recently changed to ‘living without floods’, which suggests a reorientation towards control. These findings suggest that development in the case study locations has strong path dependencies and controlling influences and interests that even a major disaster cannot disrupt. In order to be transformative, resilience-building measures need to address the drivers of risk and the causes of social vulnerability. The latter often results because of entrenched power relations that create and perpetuate unequal access to resources and can only be addressed through inclusive governance. However, power relations governing resilience to disasters are not static. Rather, since disasters open political space and can act as catalysts for change, after disasters occur, the greatest opportunity for more inclusive governance arises here (Pelling and Dill 2010). The governance of post-disaster recovery and resilience building measures was top-down in Thailand, Vietnam and Cambodia, and a combination of top-down and bottom-up in Indonesia. Only in the latter case was there an attempt to include those most affected by the disaster in the design and implementation of the recovery process. But even here, the role and support of the Panglima Laot was influenced by donors and implementing agencies, and has led to an erosion of the community culture of mutual cooperation within the Aceh fishing community. Other critical governance issues evidenced across the case studies include challenges in coordination and collaboration between different government agencies and between government agencies and other recovery actors, and transparency of decision-making processes, particularly around compensation of people and businesses for incurred losses and damages to property and livelihoods.

Conclusion In our study of longer-term recovery following major disasters in Southeast Asia, we found a diversity of loss and damage

Long-term recovery narratives following major disasters in Southeast Asia

systems which reflected the dominant recovery narratives around the disaster. The case study findings show that efforts to improve the performance of such systems so that they support long-term recovery need to closely consider the causes of vulnerability, intended beneficiaries, the framing of solutions and issues of governance such as the legitimacy, responsiveness and accountability of authorities. The findings of this study are important for improving the performance of loss and damage systems, both formal and informal, and, ultimately, supporting more climate-resilient development. The definition of disaster loss and damage systems and the typology of post-disaster recovery narratives developed in this study can be developed further to provide practical guidance for national and local decision-makers to improve policies and strategies for recovery and resilience planning and implementation. This can be done through the creation of directives under the Sendai Framework for DRR: 2015-2030 (UN General Assembly 2015), where the BBB narrative dominates, that specify principles, methods and processes for deciding what types of interventions to prioritise in different post-disaster recovery contexts and how to translate narratives into meaningful interventions that reduce the vulnerability of all people at risk. Acknowledgements The authors would like to thank the Asia Pacific Network for Global Environmental Change Research (APN) for supporting the research presented in this report. Thanks also to many individuals who helped with data collection in the field or agreed to be interviewed or provide information to the project. Special thanks, in particular, to Suttirak San-ngah, Dton Siriwan, Dr. Nguyen Duy Can, Nguyen Quynh Anh and Do Thuy Ngan.

References Audet F (2015) From disaster relief to development assistance: why simple solutions don’t work. Int J: Can J Glob Policy Anal 70(1):110– 118. https://doi.org/10.1177/0020702014562595 Birkmann J, Buckle P, Jaeger J, Pelling M, Setiadi N, Garschagen M, Fernando N, Kropp J (2010) Extreme events and disasters: a window of opportunity for change? Analysis of organizational, institutional and political changes, formal and informal responses after mega-disasters. Nat Hazards 55(3):637–655. https://doi.org/10. 1007/s11069-008-9319-2 Boyd E, James RA, Jones RG, Young HR, Otto FEL (2017) A typology of loss and damage perspectives. Nat Clim Chang 7(10):723–729. https://doi.org/10.1038/nclimate3389 Boyland M, Nugroho A, Thomalla F (2017) The role of the Panglima Laot customary institution in the 2004 Indian Ocean tsunami recovery in Aceh. In: Djalante R, Thomalla F, Garschagen M (eds) Disaster risk reduction in Indonesia: progress, challenges, and issues, disaster risk reduction. Springer, pp 357–376. https://doi.org/ 10.1007/978-3-319-54466-3_14 Cardona, O.D., Van Aalst, M.K., Birkmann, J., Fordham, M., McGregor, G., Perez, R., Pulwarty, R.S., Schipper, E.L.F., Sinh, B.T., 2012. Chapter 2: determinants of risk: exposure and vulnerability, in: Decamps, H., Keim, M. (Eds.), Managing the risks of extreme events and disasters to advance climate change adaptation. A special

report of working groups I and II of the Intergovernmental Panel on Climate Change (IPCC). International Panel on Climate Change (IPCC), Cambridge, U.K. and New York, NY, USA, pp. 65–108 Christoplos, I., 2006. Links between relief, rehabilitation and development in the tsunami response. Tsunami Evaluation Coalition, London CRED, Guha-Sapir, D., 2017. EM-DAT: the emergency events database—Université catholique de Louvain (UCL) [WWW Document]. URL http://www.emdat.be (accessed 6.7.17) Daly P (2015) Embedded wisdom or rooted problems? Aid workers’ perspectives on local social and political infrastructure in posttsunami Aceh. Disasters 39(2):232–257. https://doi.org/10.1111/ disa.12105 Fan, L., 2013. Disaster as opportunity? Building back better in Aceh, Myanmar and Haiti. Overseas Development Institute Gotham KF (2008) From 9/11 to 8/29: post-disaster recovery and rebuilding in New York and New Orleans. Soc Forces 87(2):1039–1062. https://doi.org/10.1353/sof.0.0131 Gunewardena N, Schuller M (eds) (2008) Capitalizing on catastrophe: neoliberal strategies in disaster reconstruction. AltaMira Press, Lanham, MD Hobor G (2015) New Orleans’ remarkably (un)predictable recovery: developing a theory of urban resilience. Am Behav Sci 59(10):1214– 1230. https://doi.org/10.1177/0002764215591180 Ingram JC, Franco G, Rio CR, Khazai B (2006) Post-disaster recovery dilemmas: challenges in balancing short-term and long-term needs for vulnerability reduction. Environ Sci Pol 9(7-8):607–613. https:// doi.org/10.1016/j.envsci.2006.07.006 IOB, 2013. Linking relief and development: more than old solutions for old problems? Ministry of Foreign Affairs, Netherlands James R, Otto F, Parker H, Boyd E, Cornforth R, Mitchell D, Allen M (2014) Characterizing loss and damage from climate change. Nat Clim Chang 4(11):938–939. https://doi.org/10.1038/nclimate2411 Joakim EP, Wismer SK (2015) Livelihood recovery after disaster. Dev Pract 25(3):401–418. https://doi.org/10.1080/09614524.2015. 1020764 Khasalamwa S (2009) Is Bbuild back better^ a response to vulnerability? Analysis of the post-tsunami humanitarian interventions in Sri Lanka. Norsk Geografisk Tidsskr Nor J Geogr 63(1):73–88. https://doi.org/10.1080/00291950802712152 Klein, N., 2007. The shock doctrine: the rise of disaster capitalism, 1 ed. Knopf Canada, Toronto Loewenstein, A., 2015. Disaster capitalism: making a killing out of catastrophe. Verso Books, London; New York Marks D (2015) The urban political ecology of the 2011 floods in Bangkok: the creation of uneven vulnerabilities. Pac Aff 88(3): 623–651. https://doi.org/10.5509/2015883623 Marks D, Lebel L (2016) Disaster governance and the scalar politics of incomplete decentralization: fragmented and contested responses to the 2011 floods in Central Thailand. Habitat Int Decentralizing Disaster Gov Spec Issue 52:57–66 Marks D, Thomalla F (2017) Responses to the 2011 floods in Central Thailand: perpetuating the vulnerability of small and medium enterprises? Nat Hazards 87(2):1–19. https://doi.org/10.1007/s11069017-2813-7 Mosel, I., Levine, S., 2014. Remaking the case for linking relief, rehabilitation and development. How LRRD can become a practically useful concept for assistance in difficult places. Overseas Development Institute, London Olsson P, Gunderson LH, Carpenter SR, Ryan P, Lebel L, Folke C, Holling CS (2006) Shooting the rapids: navigating transitions to adaptive governance of social-ecological systems. Ecol Soc 11:18 Pelling M, Dill K (2010) Disaster politics: tipping points for change in the adaptation of sociopolitical regimes. Prog Hum Geogr 34(1):21–37. https://doi.org/10.1177/0309132509105004

F. Thomalla et al. Sanderson D, Sharma A, Kennedy J, Burnell J (2014) Lost in transition: principles, practice and lessons from Haiti for urban post-disaster shelter recovery programs. Asian J Environ Disaster Manag 6(02): 131–151. https://doi.org/10.3850/S1793924014000362 Shanahan EA, Jones MD, McBeth MK (2011) Policy narratives and policy processes. Policy Stud J 39(3):535–561. https://doi.org/10. 1111/j.1541-0072.2011.00420.x Surminski S, Lopez A (2014) Concept of loss and damage of climate change—a new challenge for climate decision-making? A climate science perspective. Climate Dev 0(3):1–11. https://doi.org/10. 1080/17565529.2014.934770 Telford, J., Cosgrave, J., 2006. Tsunami evaluation coalition—joint evaluation of the international response to the Indian Ocean tsunami (synthesis report). Tsunami Evaluation Coalition, London Thomalla F, Larsen RK (2010) Resilience in the context of tsunami early warning systems and community disaster preparedness in the Indian Ocean Region. Environ Hazards 9(3):249–265. https://doi.org/10. 3763/ehaz.2010.0051

UN (2006) Key Propositions for Building Back Better: A Report by the UN Secretary-General’s Special Envoy for Tsunami Recovery, William J. Clinton. New York: United Nations UN General Assembly (2015) Sendai Framework for Disaster Risk Reduction 2015-2030 (No. A/CONF.224/CRP.1). UN, Sendai, Japan UNFCCC, 2015. Report of the Conference of the Parties on its twentyfirst session, held in Paris from 30 November to 13 December 2015 (No. FCCC/CP/2015/10/Add.1. 1–32). United Nations Framework Convention on Climate Change UNFCCC, 2013. Warsaw international mechanism for loss and damage associated with climate change impacts (No. Decision 2/CP.19). United National Framework Convention on Climate Change, Standing Committee on Finance UNGA, 2016. Report of the open-ended intergovernmental expert working group on indicators and terminology relating to disaster risk reduction (No. A/71/644). United Nations General Assembly World Bank (2006) Aceh Public Expenditure Analysis Spending for Reconstruction and Poverty Reduction. Washington DC: World Bank