When local response meets national narratives: reflections from the Sumatra floods

During the first week of the floods in Sumatra, many of the first images circulating online did not come from official channels. They came from residents filming rising waters with their phones, volunteers coordinating evacuation boats through messaging groups, relatives sharing updates through social media, and local organisations scrambling to mobilise food, shelter and information before formal systems had fully activated. For those of us working alongside local humanitarian actors, this moment felt both familiar and unsettling.

Familiar, because across Indonesia it is often communities and local organisations that move first when disaster strikes. Unsettling, because the scale of the disaster was not immediately recognised, with many initially perceiving the floods as part of recurring seasonal patterns. The weeks that followed revealed tensions that are less frequently discussed in global conversations about localisation. Local responders filled critical gaps in the early response, but their actions were also unfolding within a wider landscape of political narratives, institutional authority and public expectations.

Localisation is often presented as a largely technical agenda: shifting resources closer to affected communities, strengthening national systems and supporting local leadership. In principle, this reflects an important recognition that those closest to affected communities are often best positioned to respond quickly and appropriately. Yet experiences from the recent floods suggest that localisation is rarely a politically neutral space.

The assumption of neutral local actors

In recent years, localisation has become a central theme in global humanitarian discussions. Commitments under initiatives such as the Grand Bargain have emphasised strengthening national systems and supporting local leadership. Much of this conversation, however, has tended to frame localisation as a technical process focused on funding flows, partnership arrangements or capacity-strengthening.

In practice, localisation does not unfold in a vacuum. Disaster response takes place within systems of governance, public expectations and national narratives about leadership and responsibility. Local organisations may be deeply embedded in their communities, but they also operate within institutional and political contexts that shape what they can do, how visible their work becomes and how their contributions are interpreted.

This means that locally led action can carry multiple layers of risk. Local organisations often take on the operational burden of early response, mobilising quickly with limited resources, while their efforts simultaneously intersect with broader narratives about national leadership, public accountability and institutional performance. Better understanding these realities that local organisations have to navigate doesn’t undermine the localisation agenda, but helps it.

Experience from the floods: three tensions in practice

The floods in Sumatra offered a vivid illustration of how localisation unfolds in practice. In the early days of the response, local organisations, community groups and informal networks were among the first to mobilise support for affected communities. In several locations, neighbourhood associations, faith-based groups and informal solidarity networks formed the backbone of the initial response, often preceding both formal humanitarian actors and institutional coordination. They organised evacuations, distributed basic supplies and shared information in rapidly changing conditions, often with limited resources.

At the same time, local responders described the early response phase as highly fragmented. According to Syafrimet Azis, co-founder of Jemari Sakato and himself both a survivor and responder during the West Sumatra floods, many organisations operated independently because coordination mechanisms that formally existed on paper had never been fully activated in practice. Although Indonesia’s disaster management framework mandates sectoral coordination clusters for areas such as health, protection and education, in West Sumatra only the logistics cluster had previously functioned.

‘Whoever was able to move first, moved first,’ Azis reflected, describing how organisations often relied on their own networks and judgement in the absence of an operational coordination platform. As a result, local organisations frequently responded simultaneously in highly visible locations while other affected areas received little attention. Following the floods, local civil society organisations initiated discussions to activate the broader cluster system, not only for emergency response but also for preparedness and future coordination.

Haris Oematan, director of CIS Timor, is one of many local humanitarian practitioners from eastern Indonesia who previously experienced the Seroja Cyclone in 2021, similarly reflected that many responders initially underestimated the scale of the disaster because flooding in several affected areas was considered ‘normal’ seasonal flooding. Yet as rainfall intensified and reports spread across community networks, local organisations quickly recognised that this was no longer a routine event. ‘The first response came from communities,’ he said, describing how community-based organisations (i.e. neighbourhood groups, religious communities and informal solidarity networks) mobilised long before larger institutional systems fully activated.

While this demonstrated the strength and immediacy of locally rooted response, it also revealed several tensions that are less visible in broader localisation debates.

Filling the gap, bearing the cost

Local actors frequently stepped in to fill gaps in the initial response, particularly in areas where access was difficult or where official assistance had not yet reached. This role, however, was not limited to service delivery. In several affected areas, local responders described how communities relied heavily on self-help mechanisms during the early recovery phase. As Oematan noted, families often paid directly for cleaning services, debris removal and basic repairs in their own homes and communities, at considerable cost, while waiting for formal assistance to arrive. In some locations, residents organised collective clean-up efforts through local volunteer groups, while elsewhere recovery became increasingly dependent on households’ own financial capacity. While this reflects strong traditions of mutual support and self-reliance, it also raises questions about how financial and recovery risks are distributed during prolonged response periods.

For local organisations, this also meant absorbing significant operational costs and risks, including working in hazardous environments, mobilising resources without certainty of reimbursement and making rapid decisions with incomplete information. Several local organisations that had previously focused on long-term development or community assistance suddenly found themselves performing emergency humanitarian roles. As Azis observed, many groups ‘unexpectedly became emergency responders’ simply because they were already embedded in affected communities and could not ignore the scale of need around them. This blurred the boundaries between development work and humanitarian response, while increasing pressure on organisations that often lacked dedicated emergency resources or preparedness systems.

From local initiative to national narrative

As the response evolved, the visibility of local initiatives interacted with broader narratives about national leadership and response effectiveness. Community-led actions, often grounded in solidarity and mutual support, can over time become incorporated into nationally framed accounts of response. This is not necessarily the result of deliberate appropriation, but can provide a misleading picture of how severe such events are and who responds to them.

During the floods, public debate around whether the disaster should be categorised as a national disaster also shaped perceptions of the response. While local solidarity initiatives became highly visible across affected areas, discussions about national leadership, international assistance and institutional capacity increasingly dominated public narratives.

Azis noted that these debates unfolded alongside frustrations over the limited fiscal capacity of provincial authorities. At the time of the floods, regional budgets were already under pressure from broader government efficiency measures, constraining the ability of local government institutions to respond at scale. This created a complex public narrative in which provincial authorities were simultaneously criticised for moving too slowly while also operating with limited resources and insufficient national support.

In practice, this produced a situation where community-led action both filled operational gaps and reinforced broader perceptions that the disaster remained manageable within existing national systems. More broadly, it reflects how public communication, institutional reporting and political expectations shape how response efforts are understood.

These dynamics can blur distinctions between different forms of leadership and contribution, making it less clear how local agency is acknowledged, supported or sustained beyond the immediate response.

The politics of visibility

While localisation discussions often emphasise the transfer of resources and responsibilities to local actors, the distribution of risk, cost and visibility is less frequently made explicit. Local organisations and communities may carry the immediate burden of response, including operational, financial and reputational risks, while other actors retain greater control over resources, visibility or strategic direction.

These dynamics also evolve over time. While the early phase of response may see a concentration of external assistance, support often declines in the following months, even as needs persist or reemerge, particularly in contexts of recurring flooding.

Azis noted that media visibility significantly influenced the distribution of assistance. Areas that became highly visible online often received repeated waves of aid, while communities experiencing indirect losses (such as damaged farmland, disrupted livelihoods or isolated housing areas without fatalities) were frequently overlooked. In some locations, households facing long-term economic disruption did not appear in official disaster data because the damage was considered less ‘visible’ than collapsed houses or casualty figures.

These uneven patterns of visibility shaped not only where aid flowed, but also whose suffering became publicly recognised. In some districts, humanitarian presence declined significantly after the first months of response despite continued rainfall and recurring flooding. This created additional pressure on local organisations and communities to sustain response efforts with increasingly limited resources.

Local responders also reflected on how unequal visibility can unintentionally generate tensions within affected communities themselves. In some locations, repeated assistance to highly publicised areas contrasted sharply with the limited support available elsewhere, creating perceptions of imbalance and exclusion. These dynamics highlight how media attention, accessibility and public narratives can influence not only humanitarian priorities but also social relationships during recovery.

What this means for humanitarian practice

Reflecting on these experiences, the question is not whether localisation should be pursued, but how it can be understood and supported in ways that reflect the realities faced by local actors. If localisation is approached primarily as a technical shift, important dimensions of practice risk being overlooked.

Experiences from the floods suggest that local actors are often expected to absorb uncertainty and continue responding even when coordination, funding and public attention fluctuate over time. Yet these realities are not always fully reflected in international localisation discussions. Although local responders can often draw on close trusted relationships with the affected communities they come from, such proximity can wrongly be assumed to automatically translate into sustainable response capacity.

Localisation also requires a stronger emphasis on political awareness within humanitarian practice. This does not mean that humanitarian actors should become political in a partisan sense, but rather that they recognise the environments in which they operate. Understanding how narratives of national leadership are formed, how public accountability is shaped and how civic space is experienced can influence how local actors position themselves and make decisions during a response. In contexts where civic space is sensitive or contested, local actors will need to navigate additional layers of complexity that go beyond service delivery. This is particularly relevant in responses to recurring disasters, where needs extend beyond the initial emergency phase.

Such issues also raise the need for risk-sharing to be made more explicit. While local actors are often expected to respond quickly and flexibly, the risks they carry – operational, financial and reputational – are not always equally recognised or supported. More deliberate conversations between local, national and international actors about how risks are distributed, mitigated and acknowledged could help create more balanced and sustainable partnerships in practice. In contexts where responsibilities and narratives may also shift over the course of a response, maintaining clarity about roles, expectations and communication becomes particularly important.

Taken together, these considerations point towards a more nuanced view of localisation: one that goes beyond technical arrangements and engages more directly with the lived realities of humanitarian response. Experiences shared by local responders suggest that effective localisation is not only about transferring responsibility closer to affected communities, but also about understanding how responsibility, coordination and public legitimacy are unevenly distributed during crisis.

The floods in Sumatra are not unique in revealing these dynamics, but they offer a timely reminder that localisation, in practice, is shaped as much by context as by design. If localisation is to become more than a shift in language or institutional structure, humanitarian actors may need to engage more openly with the realities that local responders already navigate in practice.

Acknowledgement

This article draws on ongoing conversations and reflections with local humanitarian practitioners and civil society organisations involved in the Sumatra flood response, including partners engaged in the Humanitarian System Transformation through Local Humanitarian Leadership (HST-LHL) programme. The author is especially grateful to Syafrimet Azis of Jemari Sakato and Haris Oematan of CIS Timor for generously sharing their experiences and insights from the field.

Stella Anjani is a humanitarian practitioner working on localisation and locally led response in Indonesia, with a background in psychology and gender studies. Her experiences are supporting civil society organisations and humanitarian system transformation initiatives.