In Kaohsiung's urban resilience exercise, Sih-Wei Primary School was a food distribution centre. The site was rehearsing procedures it would follow in a crisis. I had watched county officials, volunteers, and a neighbourhood chief spend 40 minutes troubleshooting its rationing process. After each step, the officials returned to study a document on a desk and then tried again.

The neighbourhood chief understood the purpose of the exercise, as well as its limits. I told him I had noticed some of them reading what looked like an SOP earlier. Did everyone at the site have one?

He smiled. No, he said, not everyone had access to the evaluator's copy. And if a war really broke out, they wouldn't necessarily have SOPs for that either. 

His comment spoke to the reality of the exercises, which were held across Taiwan from April to August. They are a part of the whole-of-society resilience initiative, aiming to better prepare Taiwan for conflict and natural disasters and strengthen deterrence against Chinese military action. The initiative has made considerable efforts so far to build the systems, plans, and structures needed for civil resilience. As this work enters its third year, it is time for a shift in focus: towards stress-testing these plans, learning from how they fail, and making them stronger.

Regional pressures are growing, and the United States increasingly expects Taiwan to show greater self-reliance on security matters. The recent Han Kuang 42 air raid drills have been the subject of a lot of commentary. But less attention has been given to the other, arguably messier side of the work: the 城鎮韌性 exercises which bring together civilian, state, and military forces to keep essential services and infrastructure running in a crisis. 

I attended three of the exercises as an independent observer, covering 11 sites across Yunlin County, Kaohsiung City, and New Taipei City. This included emergency operations and coordination centres, food distribution sites, emergency shelters, and medical relief stations. All of my visits were conducted with the permission of local authorities and in sites which were open to public observation. I spoke to participants about their experiences and contextualised what I saw and heard against official planning and evaluation materials. I recently wrote about my observations in Yunlin. 

I came to the exercises not as a security expert, but as a practitioner who has helped to design high-risk public systems and rebuilt trust when they failed.

Overall, I saw a national programme being implemented across Taiwan with impressive consistency. But I also saw gaps which could pose risks to Taiwan’s security.

Planning documents suggested the exercises would include complex and unpredictable scenarios. But the drills I observed largely rehearsed procedures under controlled conditions, often for 15 minutes or less. Participants at Banqiao’s coordination centre tested their reporting lines with other sites. A site in Douliu practised storing rations and allocating them to volunteers. They didn’t appear to test what happens when these processes stop working: for example, when command structures break down or information is unreliable. 

One site did introduce a theoretical communications outage into their drills. But this was clearly announced up front – and from what I could see, the procedures seemed to run just as smoothly. 

Some of the relationships underpinning this work appeared to be fragile. A major study in Taiwan’s own Meizhou township indicated that the strength of local relationships can be a major factor in community resilience against natural disasters. But at the exercises I attended, I was told multiple times that many of the people working at a site had never met before that day and would not remain in contact after. There were exceptions – including at Sih-Wei Primary School, where the neighbourhood chief knew and seemed to be trusted by almost everyone on site. The group at Sih-Wei cooperated and innovated more proactively than at any other site I observed. 

The sites I observed relied heavily on the support of older, mostly retired volunteers. Very few seemed to be of working age. One neighbourhood chief I spoke to was concerned by this. He estimated that his rationing centre served around 7000 residents, and in a crisis may need to process over 1000 people at once. He needed to train volunteers who were capable of handling such numbers. But most younger people, he said, chose to go to work instead. 

Civil society and community organisations were also surprisingly scarce. Other than major groups like the Tzu-Chi Foundation, I saw little participation from NGOs or businesses. I heard about other local networks from participants, including a cluster of LINE groups that neighbourhood chiefs in Yunlin and Kaohsiung use daily to respond to issues like damaged water pipes or roads. But despite the major roles such networks can play in keeping communities functioning in a crisis, I saw no evidence that they had a formal part in the exercises.

Previous reviews have made similar findings to what I saw. A report by RAND found that compliance in earlier iterations of the exercises had been “more in form than in substance”. Despite the government stating in 2025 that drills would “start from scratch”, the German Marshall Fund noted that last year’s exercises were still criticised for being “too perfect”. In this year's exercises, the substance of these criticisms seemed to remain. Many participants I spoke to were unsure about how the evaluations would actually improve practice. 

And there is a deeper risk. ‘Whole-of-society resilience’ requires broad participation. But in Taiwan’s intense domestic political environment, civil defence itself has become increasingly contentious. If large parts of the population do not trust or support this work, then the exercises cannot create a reliable picture of how society would respond in a crisis. 

All of these challenges need to be understood as different aspects of the same problem: Taiwan’s civil defence model is not learning everything it needs to learn.

Recent analysis suggests that Ukraine's capacity for institutional learning has been one factor in its tactical successes against Russia. Taiwanese officials have themselves described societal resilience as a “continuous process of adaptation”. Putting this into practice should be a priority for the next phase of work.

But organisational learning doesn’t happen automatically – it is a question of design. Sites like Sih-Wei showed that Taiwan already has examples of high performance. These approaches to learning and improvement need to be actively planned for and rewarded across the system.

As a first step, the exercises should give people opportunities to make mistakes. These opportunities create crucial data.

This means introducing real uncertainty into their design. Scenarios should require sites to work together in conditions where intelligence is fragmented, SOPs are insufficient, and extended improvisation is necessary. This would build better crisis management skills while highlighting where Taiwan's response systems could realistically fail.

Second, create a sharper set of measuring tools.

The current evaluation framework is extensive but unwieldy. Learning and innovation only count for a fraction of the overall score a site receives. Next year, the evaluation framework should treat them as core capabilities. Rather than just verifying whether a food distribution centre has created a plan for transporting and allocating rations, it should assess how well the site was able to function after supply chains were damaged.

Third, ensure that these lessons actually change behaviour over time. 

Participants should know who is responsible for acting on identified issues and when that will happen. One way to achieve this could be to task existing structures, such as the municipal and county coordination centres, with driving and reviewing these changes across the system. To do so, these units would need authority and resources that enable them to operate beyond the exercises themselves.

Taiwan also needs to make choices about how existing community networks should be integrated into the official response.

Many of these networks, such as local civil defence associations and the neighbourhood chief LINE groups that participants described to me, hold knowledge and relationships that Taiwan should tap into. But in a hostile environment, they may also be vulnerable to outside manipulation. The exercises should test how they work with other parts of the system and directly probe them to see where they are most vulnerable. The findings should be used to keep them more effective and secure.

Outside the exercises, Taiwan needs to continue building public trust and support for whole-of-society resilience. This is critical to the long-term legitimacy of the work. 

It means building cross-party ownership wherever possible – especially at the local level. I saw evidence that this is already happening: for example, county magistrate and opposition member Chang Li-shan actively led and evaluated the exercises in Yunlin. 

And it means finding additional points of connection between communities and civil defence planners. Right now, this role is largely held by neighbourhood chiefs – but other figures, like firefighters, teachers, doctors, and civic leaders, could be supported and resourced to do so too. These kinds of relationships do not just strengthen trust; they offer alternative channels that information can flow through in a crisis. 

Taiwan is increasingly being positioned as a regional democratic leader. To succeed in this role, it must ensure that its own efforts to build whole-of-society resilience are rigorous and credible. Getting this right requires pragmatism, and honesty about what isn’t working. But this is what it means to take resilience seriously – and it is what Taiwanese people deserve.

(This piece reflects the author's opinion, and does not represent the opinion of CommonWealth Magazine.)

CommonWealth Magazine welcomes op-ed submissions. Please send your article proposals to [email protected]

About the author:

Tom Wilson is a Taipei-based governance and resilience practitioner whose work focuses on how democracies in the Asia-Pacific function under pressure. He is also an Associate Expert at the Center for National and Collective Resilience. Previously, at New Zealand's Public Service Commission, he advised Cabinet Ministers on public sector reforms, led the design of new public institutions, and developed mechanisms for cross-government coordination and partnership with Māori organisations.