Written by Tom Wilson.
Image credit: 國防部全民防衛動員署後備指揮部/ Facebook.
Taiwan’s largest ever series of urban resilience exercises began in April 2026. Running through to August, the drills are part of an ambitious push to involve all parts of society in civil defence preparations. But this work is coming up against a tension: while community actors play a crucial role in Taiwan’s crisis responses, many remain disconnected from official planning and infrastructure. If left unaddressed, this could leave a critical gap in the island’s resilience.
A widening approach to civil defence
This year’s exercises are a major set piece in Taiwan’s whole-of-society resilience initiative. As military threats from the People’s Republic of China intensify and extreme weather events become more frequent, the initiative aims to strengthen defence planning by tapping into civil society, communities, and business. The exercises will put this to the test in 11 city and county areas. Taiwan is already making progress: by February 2026, it had designated thousands of shelters and rationing stations and distributed the latest official civil defence handbook to over 9 million households.
Civil society is making its own preparations too. Alongside larger NGOs like the Tzu-Chi Foundation, which have long been an important part of Taiwan’s disaster response, organisations like Forward Alliance and Kuma Academy have begun training thousands of civilians in first aid and basic survival skills. Smaller citizens’ groups are running their own trainings and developing radio networks they can use if official communications fail. These sit alongside the many community actors – neighbourhood (里) associations, temple committees, church congregations, after-school centres – which are often first to mobilise on the ground in a crisis.
The question is how these systems should work together. Coordinating across sectors has become a major area of interest for decision-makers. President Lai Ching-te has signalled he intends to “codify” grassroots efforts to align them more closely with national planning.
But whole-of-society coordination is difficult. Taipei’s 2025 exercises showed that there is still a disconnect between official guidance and actual practice. Much of the government’s planning so far has focused on citizen training and physical infrastructure, while prioritising larger and more visible organisations. Less attention has been given to informal community actors and how they should be integrated into Taiwan’s response. This is creating a dangerous blind spot for Taiwan.
What we can learn from crises in Taiwan and overseas
Recent disasters show why these groups are needed. When Typhoon Danas struck Taiwan’s west coast on 6 July 2025, it damaged roads and cut power across the region. Many residents waited more than a week for outside assistance. In Tainan’s Houbi Township, one of the main hubs of the response was not an official civil defence centre, but a tutoring space called Puyu Academy, which had recently started delivering meals for the elderly. On the day the typhoon passed, staff prepared hundreds of meals for residents, despite its facilities being severely damaged. Puyu Academy played a central role in Houbi’s recovery. In the weeks after the storm, it became a site for distributing supplies and holding public seminars, bringing together locals and district officials to talk about how Tainan could better prepare for future emergencies.
Puyu Academy was filling a gap. Huang Ya-sheng, its founder, noted that “not every disaster can be handled by the government alone – especially when it exceeds past experience or available manpower.” In these situations, he believed it was the responsibility of citizens to step in. But the academy was also working in a constrained environment. Tainan’s mayor, Huang Wei-zhe, acknowledged there had been shortcomings in the city’s initial response and indicated that regulations would be loosened in light of the disaster, suggesting that they may have hampered residents’ ability to respond in the moment.
These patterns show up repeatedly in crisis settings. In late 2025, when floods destroyed road connections into Guangfu township, I saw firsthand how residents used LINE groups to share information and make rapid decisions about their safety. I have seen similar patterns in my own country, Aotearoa New Zealand, too. In February 2023, Cyclone Gabrielle swept thousands of tonnes of logging debris into the Mangatokerau River, inundating towns and farmland downstream. Public organisations moved quickly to help, but unclear and conflicting lines of command delayed many. In that gap, much of the early response was led by Māori communal centres called marae.
Like Puyu Academy, marae staff understood their communities and could reach households that official channels could not. But their efforts were still largely ad hoc. Staff said they were driven more by “adrenaline” than systematic planning. Later reviews found that the region’s emergency plans had not accounted for the expertise and networks that marae and other community groups held, and the government had failed to build prior trust with these groups. This slowed the response down and put lives at risk.
While working in the design of New Zealand’s public institutions, I came to think of these as ‘gentle failures’: moments where formal systems buckle under pressure and communities bear the load. These moments can be consequential in a crisis. In the first six weeks of the Ukraine war, before international donors had become active, almost all humanitarian assistance was delivered by Ukrainian NGO and volunteer networks. They could do this because they were adaptable, trusted in their communities, and had access to real-time knowledge. But without strong mechanisms connecting them to donors and the state, this work became fragmented over time – contributing to volunteer burnout and weakening the long-term response.
The tensions for Taiwan
Taiwan needs to bring these networks into its national planning without undermining the very things which make them effective.
Part of the answer lies in how resilience is measured. The government defines it as being about enabling institutions and communities to be self-sufficient in a crisis. Taiwan’s exercises measure certain elements of this: stockpile levels, supply chains, public awareness and participation. These are essential, but they overlook the relational factors which shape how resilience works in practice: how people and organisations cooperate, share information, and make decisions. Ultimately, this deprioritises informal actors and makes it harder for them to get support – even when they are vital to keeping communities running.
Taiwan’s national security environment also creates major challenges. When the very act of civil defence planning has become deeply polarised, and foreign actors increasingly target social networks, community networks become sites of risk. In a conflict setting, they are vulnerable to infiltration, can cut across official command lines, and could leak sensitive information. All of these threats show why President Lai’s goal to standardise civil defence is necessary. But controlling these groups too tightly can also weaken them. Experience in Ukraine has shown that despite the risks of foreign exploitation, the success of local networks has hinged on their ability to move quickly and make decisions on the fly.
Moving past these challenges requires a shift in thinking: community actors should be seen as expert partners, not as resources to be managed. Groups like Puyu Academy can do things the state cannot achieve on its own. The best way for the state to help them is not to create SOPs – but to give them security guarantees and the legal certainty they need to work effectively and safely.
How Taiwan can fill these gaps
The success of the next phase of this work will depend on whether officials can recognise and harness the resilience that already exists in Taiwan.
The National Science and Technology Centre for Disaster Reduction (NCDR) already has a model for this in its “disaster prevention communities,” which supports voluntary district groups to prepare for natural disaster risks. Expanding this to whole-of-society resilience could offer a way to connect communities’ own civil defence planning into national frameworks.
Second, Taiwan needs to treat trust as a security asset. While trust within Taiwanese communities tends to be high – as shown in the 2024 Hualien earthquake and last year’s Guangfu floods – the trust between citizens and institutions is mixed. Officials should track these as key indicators of societal resilience. Where trust has broken down, the state needs to invest in it. One way to do this is to create liaison roles that build and maintain connections between communities and institutions in the long term – not just during drills and exercises.
Finally, Taiwan should continue seeking real-world data on how community resilience operates in conflict and disaster settings. Many countries, including New Zealand, prepare post-disaster governance reviews which examine how people and organisations worked together in the response. Taiwan should draw on these and conduct its own, complementing NCDR’s existing technical reviews, to better understand these dynamics in past disasters. This will be indispensable for planning better emergency responses in the future.
Taiwan has put the building blocks in place to make its society safer and stronger. The next question is about coordination: how can institutions and communities best support one another in a crisis? This is not an easy task – but ignoring it will leave a major hole in Taiwan’s resilience.
Tom Wilson is a public governance practitioner with over a decade of experience working across government and civil society in New Zealand and Taiwan. He has led the design of new public institutions, and has developed mechanisms for cross-government coordination and for involving communities and Māori tribal authorities in public decision-making. He is currently based in Taipei, where his work focuses on strengthening democratic resilience in the Asia-Pacific region.